bijournal.hse.ru 13-1-2019 e… · hse scientific journal publisher: national research university...

HSE SCIENTIFIC JOURNAL

Publisher:

National Research University

Higher School of Economics

Subscription index

in the Rospechat catalog –

80870

The journal is published quarterly

The journal is included into the list of peer reviewed scientific editions established by the Supreme Certification

Commission of the Russian Federation

Editor-in-Chief:

A. Golosov

Deputy Editor-in-Chief

S. Maltseva

Y. Koucheryavy

Computer Making-up:

O. Bogdanovich

Website Administration:

I. Khrustaleva

Address:

33, Kirpichnaya Street, Moscow,

105187, Russian Federation

Tel./fax: +7 (495) 771-32-38

http://bijournal.hse.ru

E-mail: [email protected]

Circulation: English version – 300 copies, Russian version – 300 copies,

online versions in English and Russian – open access

Printed in HSE Printing House

3, Kochnovsky Proezd, Moscow,

Russian Federation

© National Research University

Higher School of Economics

Vol. 13 No 1 – 2019

Information systems and technologies in businessV.I. Ananyin, K.V. Zimin, R.D. Gimranov, M.I. Lugachev, K.G. Skripkin

Real time enterprise management in the digitalization era ............. 7

Data analysis and intelligence systemsA.V. Demidovskij, E.A. Babkin

Developing a distributed linguistic decision making system ........... 18

Modeling of social and economic systemsA.S. Akopov, A.L. Beklaryan, M. Thakur, B.D. Verma

Developing parallel real-coded genetic algorithms

for decision-making systems of socio-ecological

and economic planning ................................................................ 33

M.A. Myznikova, L.N. Brazhnikova

Development of strategic management tools

for heat supply enterprises in the Donetsk region .......................... 45

N.K. Khachatryan, G.L. Beklaryan, S.V. Borisova, F.A. Belousov

Research into the dynamics of railway track capacities

in a model for organizing cargo transportation between

two node stations ......................................................................... 59

Information securityM.V. Tumbinskaya, B.I. Bayanov, R.Zh. Rakhimov, N.V. Kormiltcev, A.D. Uvarov

Analysis and forecast of undesirable cloud services traffic ............. 71

BUSINESS INFORMATICS Vol. 13 No 1 – 2019

2

ABOUT THE JOURNAL

Business Informatics is a peer reviewed interdisciplinary academic journal published since

2007 by National Research University Higher School of Economics (HSE), Moscow,

Russian Federation. The journal is administered by School of Business Informatics.

The journal is published quarterly.

The mission of the journal is to develop business informatics as a new field within both information

technologies and management. It provides dissemination of latest technical and methodological

developments, promotes new competences and provides a framework for discussion in the field of

application of modern IT solutions in business, management and economics.

The journal publishes papers in the areas of, but not limited to:

data analysis and intelligence systems

information systems and technologies in business

mathematical methods and algorithms of business informatics

software engineering

internet technologies

business processes modeling and analysis

standardization, certification, quality, innovations

legal aspects of business informatics

decision making and business intelligence

modeling of social and economic systems

information security.

The journal is included into the list of peer reviewed scientific editions established by the Supreme

Certification Commission of the Russian Federation.

The journal is included into Web of Science Emerging Sources Citation Index (WoS ESCI) and

Russian Science Citation Index on the Web of Science platform (RSCI).

International Standard Serial Number (ISSN): 2587-814X (in English), 1998-0663 (in Russian).

Editor-in-Chief: Dr. Alexey Golosov – President of FORS Development Center, Moscow,

Russian Federation.

EDITOR-IN-CHIEF

Alexey Golosov FORS Development Center, Moscow, Russia

DEPUTY EDITOR-IN-CHIEF

Svetlana Maltseva National Research University Higher School of Economics, Moscow, Russia

Yevgeni Koucheryavy Tampere University of Technology, Tampere, Finland

EDITORIAL BOARD

Habib Abdulrab National Institute of Applied Sciences, Rouen, France

Sergey Avdoshin National Research University Higher School of Economics, Moscow, Russia

Andranik Akopov National Research University Higher School of Economics, Moscow, Russia

Fuad Aleskerov National Research University Higher School of Economics, Moscow, Russia

Alexander Afanasyev Institute for Information Transmission Problems (Kharkevich Institute), Russian Academy of Sciences, Moscow, Russia

Anton Afanasyev Central Economics and Mathematics Institute, Russian Academy of Sciences, Moscow, Russia

Eduard Babkin National Research University Higher School of Economics, Nizhny Novgorod, Russia

Sergey Balandin Finnish-Russian University Cooperation in Telecommunications (FRUCT), Helsinki, Finland

Vladimir BarakhninInstitute of Computational Technologies, Siberian Branch of Russian Academy of Sciences, Novosibirsk, Russia

Alexander Baranov Federal Tax Service, Moscow, Russia

Jorg BeckerUniversity of Munster, Munster, Germany

Vladimir Belov Ryazan State Radio Engineering University, Ryazan, Russia

Alexander Chkhartishvili V.A. Trapeznikov Institute of Control Sciences, Russian Academy of Sciences, Moscow, Russia

Vladimir Efimushkin Central Research Institute of Communications, Moscow, Russia

Tatiana Gavrilova Saint-Petersburg University, St. Petersburg, Russia

Herv GlotinUniversity of Toulon, La Garde, France

Andrey Gribov CyberPlat Company, Moscow, Russia

Alexander Gromoff National Research University Higher School of Economics, Moscow, Russia

Vladimir Gurvich Rutgers, The State University of New Jersey, Rutgers, USA

Laurence Jacobs University of Zurich, Zurich, Switzerland

Liliya Demidova Ryazan State Radio Engineering University, Ryazan, Russia

EDITORIAL BOARD

Iosif Diskin Russian Public Opinion Research Center, Moscow, Russia

Nikolay Ilyin Federal Security Guard of the Russian Federation, Moscow, Russia

Dmitry Isaev National Research University Higher School of Economics, Moscow, Russia

Alexander Ivannikov Institute for Design Problems in Microelectronics, Russian Academy of Sciences, Moscow, Russia

Valery Kalyagin National Research University Higher School of Economics, Nizhny Novgorod, Russia

Tatiana Kravchenko National Research University Higher School of Economics, Moscow, Russia

Sergei Kuznetsov National Research University Higher School of Economics, Moscow, Russia

Kwei-Jay LinNagoya Institute of Technology, Nagoya, Japan

Mikhail Lugachev Lomonosov Moscow State University, Moscow, Russia

Peter Major UN Commission on Science and Technology for Development, Geneva, Switzerland

Boris Mirkin National Research University Higher School of Economics, Moscow, Russia

Vadim Mottl Tula State University, Tula, Russia

Dmitry Nazarov Ural State University of Economics, Ekaterinburg, Russia

Dmitry Palchunov Novosibirsk State University, Novosibirsk, Russia

Panagote (Panos) Pardalos University of Florida, Gainesville, USA

scar PastorPolytechnic University of Valencia, Valencia, Spain

Joachim Posegga University of Passau, Passau, Germany

Kurt Sandkuhl University of Rostock, Rostock, Germany

Yuriy Shmidt Far Eastern Federal University, Vladivostok, Russia

Christine Strauss University of Vienna, Vienna, Austria

Ali Sunyaev Karlsruhe Institute of Technology, Karlsruhe, Germany

Victor Taratukhin University of Munster, Munster, Germany

Jos TriboletUniversidade de Lisboa, Lisbon, Portugal

Olga Tsukanova Saint-Petersburg National Research University of Information Technologies, Mechanics and Optics, St. Petersburg, Russia

Mikhail Ulyanov V.A. Trapeznikov Institute of Control Sciences, Russian Academy of Sciences, Moscow, Russia

Raissa Uskenbayeva International Information Technology University, Almaty, Kazakhstan

Marcus Westner Regensburg University of Applied Sciences, Regensburg, Germany


4

ABOUT THE HIGHER SCHOOLOF ECONOMICS

Consistently ranked as one of Russia’s top universities, the Higher School of

Economics (HSE) is a leader in Russian education and one of the preeminent

economics and social sciences universities in Eastern Europe and Eurasia.

Having rapidly grown into a well-renowned research university over two decades, HSE

sets itself apart with its international presence and cooperation.

Our faculty, researchers, and students represent over 50 countries, and are dedicated

to maintaining the highest academic standards. Our newly adopted structural reforms

support both HSE’s drive to internationalize and the groundbreaking research of our

faculty, researchers, and students.

Now a dynamic university with four campuses, HSE is a leader in combining Russian

educational traditions with the best international teaching and research practices. HSE

offers outstanding educational programs from secondary school to doctoral studies,

with top departments and research centers in a number of international fields.

Since 2013, HSE has been a member of the 5-100 Russian Academic Excellence

Project, a highly selective government program aimed at boosting the international

competitiveness of Russian universities.


5

ABOUT THE SCHOOL OF BUSINESS INFORMATICS

The School of Business Informatics is one of the leading divisions of HSE’s

Faculty of Business and Management. The School offers students diverse courses

taught by full-time HSE instructors and invited business practitioners. Students

are also given the opportunity to carry out fundamental and applied projects at various

academic centers and laboratories.

Within the undergraduate program, students participate each year in different case-

competitions (PWC, E&Y, Deloitte, Cisco, Google, CIMA, Microsoft Imagine CUP,

IBM Smarter Planet, GMC etc.) and some of them are usually as being best students by

IBM, Microsoft, SAP, etc. Students also have an opportunity to participate in exchange

programs with the University of Passau, the University of Munster, the University of

Business and Economics in Vienna, the Seoul National University of Science and

Technology, the Radbound University Nijmegen and various summer schools (Hong

Kong, Israel etc.). Graduates successfully continue their studies in Russia and abroad,

start their own businesses and are employed in high-skilled positions in IT companies.

There are four graduate programs provided by the School:

Business Informatics

E-Business;

Information Security Management;

Big Data Systems.

The School’s activities are aimed at achieving greater integration into the global

education and research community. A member of the European Research Center for

Information Systems (ERCIS), the School cooperates with leading universities and

research institutions around the world through academic exchange programs and

participation in international educational and research projects.


7

Real time enterprise management in the digitalization era

Vladimir I. Ananyin a


Konstantin V. Zimin b E-mail: [email protected]

Rinat D. Gimranov c E-mail: [email protected]

Mikhail I. Lugachev d E-mail: [email protected]

Kirill G. Skripkin d E-mail: [email protected]

a Russian Presidential Academy of National Economy and Public Administration Address: 82, Prospect Vernadskogo, Moscow 119571, Russia

b The Russian Union of CIO Address: 34, Seleznevskaya Street, Moscow 123056, Russia

c PJSC Surgutneftegaz Address: 1 block 1, Grigoriya Kukuevitskogo Street, Surgut 628415, Russia

d Lomonosov Moscow State UniversityAddress: 1 build. 46, GSP-1, Leninskie Gory, Moscow 119991, Russia

Abstract

This paper discusses real time control of an enterprise. The history of this concept is associated with the arrival of the real time enterprise (RTE) concept in 2002. The RTE concept has been interpreted variously, mainly in the areas of computer science and marketing. With the advent of new digital technologies and digital organizations, the RTE concept has received a new practical application in management.

This paper discusses an important characteristic of the RTE concept – real time scale and the division value of this scale. The authors have investigated the factors infl uencing the division value of this scale. The composition of these factors includes not only management, but also digitalization factors. We propose considering the real time scale as a time characteristic of organization adaptation to dynamics, uncertainties and complexities that are present in its environment. In this case, the division value of the real time scale is the time that characterizes the limit after which there is a loss of control in the organization.

INFORMATION SYSTEMS AND TECHNOLOGIES IN BUSINESS


8

There are two groups of factors infl uencing the division value of the real time scale: objective factors (for example, the speed of the actual operating processes), and subjective factors (for example, limitations on participants’ knowledge of the real situation and/or their opportunistic behavior). Nevertheless, the real time scale is a real phenomenon which has objective manifestations. In a well managed organization, management always spontaneously reaches a consensus regarding the division value of the real time scale. Meanwhile, the division value of real time scale is the time division value of a real clock which is suffi cient for precise planning and control of deviations from the plan.

Key words: digital enterprise; real time enterprise; variability; enterprise manageability; dynamics,

uncertainties and complexities of environment.

Citation: Ananyin V.I., Zimin K.V., Gimranov R.D., Lugachev M.I., Skriprin K.G. (2019)

Real time enterprise management in the digitalization era. Business Informatics, vol. 13, no 1,

pp. 7–17.

DOI: 10.17323/1998-0663.2019.1.7.17

Introduction

Due to digitalization, a “technologi-

cal rearmament race” has already

begun, and it is accelerating. Its

main goal is not just to introduce new infor-

mation technologies but to digitize businesses

as well. It is shown in [1] that digitalization

of enterprises creates innovative management

practices in the fields of organizational, infor-

mational, and human capital. These new prac-

tices are complementary, and they are mutually

enhancing each other.

Among these new practices, the most impor-

tant is real time management. We have to note

that real time management is not a completely

new trend in traditional management. Indeed,

the jobs of a manufacturing process operator

or a railway freight dispatcher are examples

of well-studied practices in real time manage-

ment. When processes are stable, the response

time of an operator or dispatcher must ensure

the process’ continuity (i.e., the manufactur-

ing must be maintained at a constant pace, or

trains must move at a certain average speed).

In these cases, the procedure of real time man-

agement is determined by the speed of the pro-

cess.

How does digitalization change the concept

of real time management? As a result of digital-

ization, the manufacturing processes or railway

freight market conditions may be constantly

changing “on the fly.” No longer can we define

the change as a transition from one stable state

to another. Digitalization gives us an opportu-

nity to get a lot of new data on a manufacturing

process or the state of a railway freight system,

and we will be able to change everything on the

fly as well. As a result, we will rarely consider

the situation as a stable one; far from it, sta-

ble states may become exceptions rather than

regular practice. Moreover, the changes them-

selves are transformed and become less pre-

dictable. For example, a railway traffic jam



9

used to be a threat for a freight dispatcher, and

it caused delays and fines. However, when the

dispatcher has more information, he has new

options: first, there can be new clients requir-

ing new routes; second, instead of a client’s

own trains, the dispatcher might be able to use

a competitor’s empty trains stuck in the jam

on the same route. While this unique situa-

tion lasts, all the participants have to quickly

agree, act on, and profit from it. Therefore, in

the times of digitalization we have to deal with

a stream of unique managerial situations rather

than with regular processes.

The decision (a collaborative one!) must be

made as quickly as it was made before by a dis-

patcher alone. What determines the procedure

of RT management in this case? what does it

depend on? how is it related to digitalization?

Those are the questions we will try to answer in

this paper.

1. History of the real time

management concept

The real time management of an enterprise

(or its separate entities) is the most impor-

tant feature in a digital organization. The real

time enterprise (RTE) concept has a long and

rich history. This concept has already been dis-

cussed for some time, but in October 2002 it was

clearly defined for the first time by Gartner’s

analysts [2]. According to this definition, an

RTE is an enterprise that competes by using

up-to date information to progressively remove

delays in the management and execution of its

critical business processes.

There are three important elements in this

definition:

1. RTE is a relatively abstract objective to

strive towards rather than a particular state of

an enterprise. As Gartner’s analysts noted [2],

“It is unlikely that an enterprise will declare

itself to have become “an RTE”... Progres-

sion is asymptotic — real world organizations

will always remain inefficient in their speed of

response… Optimal RTE capability is a mov-

ing target…” In this concept, real time criteria

must be relative and varying;

2. Information is necessary but not sufficient.

Using up-to date information, we can move

towards the target (RTE), but we will be need-

ing more than just the information, because its

use requires actions and other assets as well.

Analysis of RTE’s activities should be based

not only on computer capital assets, but on

other complementary assets as well [1];

3. Gartner’s experts [2] distinguished two

areas where the RTE concept may be used: exe-

cution of operational processes and activities

management. They note that at the beginning,

enterprises were mostly focusing on the oper-

ational processes on their way towards RTE.

However, application of the RTE concepts

to the expert activities of knowledge workers,

as well as to management problems, could be

beneficial. Therefore, Gartner’s analysts state

that RTE can be used under circumstances of

both a routine issue and an emergency.

The RTE idea was welcomed by many organ-

izations and experts. We can identify the fol-

lowing two interpretations of RTE: informa-

tional and managerial.

Informational interpretation of RTE. The

RTE concept was first used by IT solution pro-

viders [3–16]. However, their understanding

of the RTE concept was limited. They defined

RTE as an organization that collects up-to-

date data and provides the necessary informa-

tion in real time to its employees, clients, sup-

pliers and collaborators. In other words, all the

information an enterprise possesses is real time

information. Usually, supporters of this inter-

pretation of RTE claim that this would happen

when manual labor is kept to a minimum, and

processes are fully automated. It is rather obvi-

ous that the informational RTE can be reached

only by a large-scale deployment of informa-

tional technologies. However, the advocates of

the informational interpretation of RTE do not

go beyond this rather obvious idea.



10

Managerial interpretation of RTE. Some

experts and organizations interpret the RTE

idea in a wider sense, claiming that the whole

cycle from making decisions to responding to

changes should function in real time [17–19].

They define RTE as an enterprise that detects

changes in operational and business conditions

and ensures a quick response to those changes.

Here, RT operation is assumed over the entire

management cycle, from capturing an event

(incident), to its analysis and decision-mak-

ing, to a response action. In addition to real

time acquisition of the information about cur-

rent events, the real time concept relies on two

more crucial stages. First, the decisions must

be “real,” i.e., we have to analyze the informa-

tion, understand the consequences, and work

out the response - all of this in real time. Sec-

ond, the proposed actions and activities should

be “real,” using and adapting the existing pro-

cesses and practices in real time. Therefore, the

real time mode must be supported by the entire

infrastructure, processes, assets and company

employees. This interpretation of real time is

deeper and closer to practical activities than

the informational interpretation.

Let us note that there are other interpreta-

tions of RTE, but they are relatively scarce and

not that important. For instance, one of them

considers RTE as a concept that gathers the

majority of new managerial ideas: information

management, big data management, knowl-

edge management, mobile enterprise, social

enterprise, etc. In our opinion, such an exces-

sive extension of the RTE concept is unjusti-

fied and impractical.

2. The RTE concept

All the experts agree that we have just begun

to study the concept of real time in the RTE

concept. As shown in [1], digitalization can

make an enterprise very competitive. However,

to take advantage of this possibility, the man-

agers of all levels as well as the employees must

make the “right” decisions. This means:

the decisions must obey a certain set of

requirements that satisfy both the solution

developers and customers; such decisions must

be implementable;

the decisions must be timely;

the decisions must be cost-effective: in

their implementation, the management sys-

tem must account for the costs of coordination

between the decision makers and participants.

The coordination costs can be calculated as the

number of man-hours that participants with a

certain level of proficiency spent to make and

implement the decision. These costs are sim-

ilar in nature to transaction costs in manage-

ment [20, 21].

Digitalization provides powerful tools for

making high-quality decisions and makes it

possible to drastically decrease the coordina-

tion costs. Digitalization also helps to make

decisions in a mode close to real time. How-

ever, even if decisions are made faster than

before, this does not mean that they are neces-

sarily timely. This problem is especially impor-

tant when an organization and/or its external

conditions are highly volatile. Let us discuss

what is real time, and how this concept is con-

nected with timeliness of the decisions.

The concept of real time characterizes a pro-

cess of management resolution in an organi-

zation, i.e., when an event requires a mana-

gerial response. As we mentioned above, such

events may constitute either a routine issue or

an emergency. It takes time to resolve such a

situation, and the amount of time should be

appropriate to prevent a routine situation from

becoming an emergency, or an emergency from

becoming a crisis or even a catastrophe:

,

where – real time of the management resolu-

tion cycle;

– the acceptable timeframe for the resolu-

tion of the managerial issue.

Note that we are talking here not only about



11

acquiring and processing the information and

decision-making; we consider the complete

cycle of resolving a managerial emergency,

which includes, apart from the steps mentioned

above, the implementation of the decisions as

well as the corresponding changes. Therefore,

we think that there is a need to study the man-

agerial interpretation of the real time manage-

ment concept more thoroughly as a complete

real time resolution of a managerial incident at

an enterprise.

When conditions are stable and predicta-

ble, the majority of managerial situations can

be easily resolved through the following steps:

information acquisition – information classi-

fication – known solution – quick response.

When conditions are unique and unpredicta-

ble, more complicated and coordinated actions

of the participants are required to resolve a

managerial issue: information acquisition –

situation evaluation – looking for and making

a coordinated decision – organization of and

control over the decision implementation. In

the latter case, all the participants must plan

their actions, hence each participant that con-

tributes to the decision has its own time scale.

The management time scale is a certain series

of time unit intervals that determine the detal-

ization level (quants) of planning and control

over the activities aimed at the resolution of

a managerial situation. A single-unit interval

should be determined by the dynamics of the

development of the managerial situation, i.e.,

by the acceptable timeframe ( ), so that the sit-

uation will not develop into a crisis or a catas-

trophe.

3. Factors that affect

the management time scale

Let us try to figure out the factors that influ-

ence a management time scale’s unit interval.

The more complex a managerial situation is

for its participants, the more complicated is the

activity aimed at its resolution. Hence, a real

managerial situation resolution cycle ( ) has

to contain more actions, while the acceptable

timeframe ( ) is fixed. Therefore, the manage-

ment time scale unit interval will be smaller in

this case. The difficulty of a managerial situ-

ation is always determined by its participants;

therefore, it has both objective and subjective

components. For the sake of simplicity, we can

state that the difficulty of a managerial situa-

tion is determined by the following four key

factors:

Scale. The difficulty of a managerial situa-

tion can depend on its scale, when there are

many interconnected factors to consider. In

this case, the time needed to find the solutions

and to resolve the situation ( ) is hardly pre-

dictable;

Information. A situation may be deemed

complicated because the participants do not

possess complete, reliable, or up-to-date infor-

mation. They will have to look for additional

information, and it is hard to predict how much

time this would take. In reality, this means that

the information should be found as quickly as

possible, and the management time scale unit

interval should be minimal;

Human capital. The situation may be consid-

ered to be complex either because it is unique,

or the personnel have never faced this situation

before (no personal experience), or they do not

know who has such an experience (for exam-

ple, they do not know that a competitor has

had such an experience, or that organization is

not willing to share it). In such situations, the

participants might “reinvent the wheel” by trial

and error, meaning that their possibilities to

plan the activities aimed at the managerial sit-

uation resolution will be very limited. In reality,

this means that the problem should be resolved

as quickly as possible, and the management

time scale unit interval should be minimal;

Organizational capital. A situation may

be difficult because the participants are not

authorized to resolve it, and escalation or del-

egation mechanisms do not work. The lack of



12

such skills as standard practices of team work,

meaningful task formulation, planning, con-

trol, and effective communications can sub-

stantially complicate the situation. In this sce-

nario, it is hard to manage the situation, once

again meaning that the management time scale

unit interval should be kept to a minimum

(everything should be done as quickly as pos-

sible to create a time leeway).

When the acceptable timeframe to resolve

a managerial situation ( ) can be decreased,

this decreases the management time scale

unit interval as well. A managerial situation is

always caused by some key reason. Most often

such reasons arise outside the organization as

impending threats or new opportunities caused

by external processes with their own dynam-

ics. In the first approximation, such dynam-

ics can be described by four key characteristics.

The first one is related to the regular course of

a process, with the remaining three reflecting

its volatility:

the speed of an external process (produc-

tivity);

the extent of variation in the external pro-

cess during a period of time (variations in the

whole process or in its subprocesses);

the number of variations in the external

process during a period of time (two variations

in the whole process or 50 variations in some

subprocesses per year);

the average speed of these subprocesses in

the external process (variations in the scope of

the whole process take three months on aver-

age; variations in the scope of a subprocess take

about a week).

It is noteworthy that key factors causing a

managerial issue may reveal themselves inside

an organization as well, for example as busi-

ness innovations or management initiatives,

with no apparent external changes. However,

they could also be characterized by the varia-

bility parameters discussed above.

In response to external changes, the enter-

prise management makes certain decisions1. We

can obtain an estimate of the acceptable time-

frame ( ) to resolve the situation in the scope

of this managerial decision. This estimate has

both objective and subjective components. In

reality, the acceptable timeframe of the situa-

tion resolution is usually decreased (this can be

described by the catchphrase “this should have

been done yesterday”). This is caused by three

main factors:

1. The increase in the speed of external pro-

cesses, their volatility, and the growth of inno-

vational activity within the organization itself

objectively require resolving any manage-

rial issues faster, meaning that the acceptable

timeframe ( ) to resolve the issue should be

decreased;

2. The acceptable timeframe ( ) to resolve a

managerial situation is decreased because of

the uncertainty in the evaluation of the situa-

tion. The participants in a managerial situation

may not possess enough knowledge or infor-

mation to correctly evaluate the scope and dif-

ficulty of the factors that caused the situation.

In this case, the participants will have to over-

estimate the required time to have a margin of

security, thus decreasing the acceptable time-

frame ( );

3. An uncertainty in the evaluation of a man-

agerial situation (for instance, an underes-

timated scale of a disaster) leads to errone-

ous estimates of the acceptable timeframe

( ). When the participants realize their mis-

take, they will need more time to correct them,

and the situation’s resolution will occur under

stricter time constraints, thus narrowing the

acceptable timeframe ( ).

For mature businesses under professional

management, emergency situations should

be rare. The majority of managerial activities

are related to routine situations, which have a

local scale, their causation is well-known, the 1 We assume that ignoring the situation and not taking action at all is also a managerial decision



13

experience of their resolution has been accu-

mulated, and all the necessary information can

be found quickly. Based on a stream of rou-

tine managerial situations, such organizations

set rules, regulations, standards, and organiza-

tional structures. In particular, they set a par-

ticular timeframe to resolve a routine manage-

rial situation, thus setting the time management

unit interval.

Unlike emergency managerial situations, the

routine ones do not cause a strong pressure to

decrease the management time scale unit inter-

val. Nevertheless, we have to remember that all

the managerial situations in an organization

are intertwined; it is hard to predict which situ-

ation might escalate or defuse; a routine situa-

tion may become an emergency, and vice versa.

Therefore, the management time scale unit

intervals must be constant for the entire stream.

Since a management system should always be

ready for emergency events, the management

time scale unit interval should be determined

by the resolution cycle ( ) of the most compli-

cated situation the organization has ever dealt

with. This does not mean that routine issues

must be resolved at the same speed as emer-

gency situations. Of course, different mana-

gerial situations should have different time-

frames. However, we think that a system that

manages an entire stream of events should have

the same timescale, and the resolution of all

the situations should be planned accounting

for the management time scale unit intervals.

This leads us to the definition of the real time

management scale.

4. The real time management scale

The real time management scale is a scale

where a single-unit interval is sufficient to

resolve the most complicated managerial sit-

uation the organization has ever dealt with. A

single-unit interval on this scale is determined

as the time necessary for the resolution of this

managerial situation ( ) divided by the number

of stages in the resolution cycle.

Sometimes a complicated managerial situa-

tion can be resolved easily and elegantly. How-

ever, this does not mean that the unit inter-

val of the real time management scale must

be increased. First, the real time management

scale describes the entire stream of managerial

situations. Therefore, for this elegant solution

to increase the unit interval on the scale, such

elegant solutions must become a regular man-

agement practice. Second, when the solution

has not been found yet, a good manager should

base any decisions on the most pessimistic sce-

nario.

We can say that an organization resolves all

the managerial situations they are aware of in a

timely manner when a routine situation never

becomes a crisis. Therefore, this scale reflects

the organization’s real time. In this context

“real” means that the time corresponds to a

certain external reality, is appropriate to the

environment, and reflects the external condi-

tions. We can define the term real time only

in the context of a link between the process

of object management and the object’s envi-

ronment. We have to emphasize that this only

concerns the environment known to the man-

agement and the external conditions they are

aware of. This relates to the note mentioned

above that the real time criteria are relative

and volatile. It is impossible to develop an

real time management scale in advance for

all unknown future managerial situations. Of

course, the unit interval of an real time man-

agement scale can be decreased proactively to

respond to more complicated managerial sit-

uations. However, there is no way of knowing

if this response would correspond to the real

time criteria.

We can say that the organization’s real time

management scale is a time parameter that

reflects how the organization adapts to the

dynamics and complexity of its environment.

A unit interval on this scale corresponds to the

limit where the management starts to lose con-

trol over the organization.



14

5. The real time scale and digitalization

Classical automation of regular stable busi-

ness processes helped to make the majority of

routine situations standard and exclude them

from the overall stream of managerial situa-

tions. This created conditions for the manage-

ment to reduce the timeframe of the resolu-

tion cycle ( ). However, this timeframe was not

reduced substantially, because the actual prac-

tice of the implementation of the resolution

decisions – even though the decisions them-

selves were made faster – remained the same;

also, strong incentives to decrease the time-

frame of the management resolution cycle ( )

were scarce. This is why classical automation

had practically no effect on the real time man-

agement scale. Effects were noticeable when

automation led to a productivity increase in

standard operational processes.

Digitalization drastically changes the stream

of managerial situations and gives the manage-

ment strong incentives to substantially reduce

the acceptable timeframe of resolution of man-

agerial situations ( ).

Increase in the density of the stream of inci-

dents and emergency situations. As digitaliza-

tion expands, the number of digital twins of real

objects increases. These twins serve as big data

sources. The data appear as soon as an event is

automatically registered, and the data volume

increases substantially. The sensitivity of an

enterprise’s management to external changes

grows; as the volume of information increases,

the participants can see risks and possibilities

they have not seen before. Now they need to

adequately respond to them ( ). This means

that the number of managerial issues, as well

as the percentage of complicated emergency

situations, will increase; this calls for the unit

interval of the RT management scale to be

decreased.

Increased complexity of an enterprise infor-

mational model. As data volume increases,

digitalization provides the participants with

efficient tools for intellectual analytics, which

allows them to find new connections and

trends. However, these new connections and

trends can only be revealed if the participants

improve their skills and use more complicated

decision-making models. For example, when

a switch is made from the business processes

scale to the scale of value chains, all the par-

ticipants must update their way of thinking to

embrace this new scale. In this case, the num-

ber of participants in managerial situations will

increase, and the situations will become more

complicated. The growth in complexity and

uncertainty once again leads to a smaller value

of the acceptable timeframe ( ) of managerial

situations.

The increase in the number of internal initi-

atives on changes. As digitalization expands,

more participants will be involved into activ-

ities related to business innovations or man-

agement initiatives. This means that the vola-

tility initiated by the organization or a value

chain must increase. The more local the ini-

tiative is, the easier it is to manage it, and the

closer it will be to a routine managerial situ-

ation. We have to comprehend that new ini-

tiatives on different scales will join the gen-

eral stream of managerial situations. This will

expand the stream and can potentially lead to

an increase in the number of complex emer-

gency situations due to the complexity of the

connections between the elements. Again,

there is a trend to decrease both the accept-

able timeframe ( ) and the unit interval of the

real time management scale.

Escalation of market competition. The phase

when the digitalization leaders on the market

are “skimming cream” will be short. Solution

developers and consultants will quickly intro-

duce new technologies to the competitors. This

will lead to a management “arms race” aimed

at decreasing the market-average values of

and . This, in turn, will create a new power-

ful incentive to decrease the unit interval of the

RT management scale.



15

Digital organizations are just starting to

appear, and we can assume that, as the digitali-

zation scope and depth expand, the unit inter-

val of the real time management scale will be

decreased. This means that when the digital

economy becomes a reality the operational and

managerial processes, as well as the pace of life,

will accelerate.

Conclusion

The real time management scale is a char-

acteristic time parameter that shows how an

organization adapts to the dynamics and com-

plexity of its environment. The unit interval of

the real time management scale sets limits to

the unit intervals of incoming signals and the

acceptable lag of a response to changes. There-

fore, the unit interval of the real time manage-

ment scale determines the limits of the possi-

bilities of managing situations.

Figuratively speaking, under classical auto-

mation, the clock in the central control room

of an enterprise reflects the real manage-

ment time of manufacturing processes. In a

digital enterprise, the real management time

is reflected by the clocks in the negotiation

rooms where decisions are made. The negoti-

ation rooms may be real as well as virtual. It is

of utmost importance though that the clocks in

those rooms be synchronized.

We can assume that the expansion of digitali-

zation in scope and depth will cause a decrease

of the unit time interval of the real time man-

agement scale.

References

1. Ananyin V.I., Zimin K.V., Lugachev M.I., Gimranov R.D., Skriprin K.G. (2018) Digital

organization: Transformation into the new reality. Business Informatics, no 2, pp. 45–54.

2. Drobik A., Raskino M., Flint D., Austin T., MacDonald N., McGee K. (2002) The Gartner

definition of real-time enterprise. Available at: https://www.gartner.com/doc/372176/gartner-

definition-realtime-enterprise (accessed 01 October 2018).

3. Intel Corporation (2014) Real-time big data analytics for the enterprise. White paper. Available

at: https://www.intel.com/content/dam/www/public/us/en/documents/white-papers/big-

data-hadoop-real-time-analytics-for-the-enterprise-paper.pdf (accessed 01 October 2018).

4. Infosys Limited (2018) Architecting the real-time enterprise. Available at: https://www.infosys.

com/business-application-services/white-papers/Documents/architecting-real-time-enter-

prise.pdf (accessed 01 October 2018).

5. SAP (2012) Surgutneftegas takes HANA for a test drive. SAP Insider Profiles.

6. Stonebraker M., etintemel U. (2005) One size fits all: An idea whose time has come and gone.

Proceedings of the 21st International Conference on Data Engineering (ICDE 05), Tokyo, Japan,

5–8 April 2005, pp. 2–11.

7. Stonebraker M., Madden S., Abadi D.J., Harizopoulos S., Hachem N., Helland P. (2007)

The end of an architectural era (It’s time for a complete rewrite). Proceedings of the 33rd

International Conference on Very Large Data Bases (VLDB 2007), Vienna, Austria,

23–28 September 2007, pp. 1150–1160.

8. Plattner H. (2009) A common database approach for OLTP and OLAP using an in-memory

column database. Proceedings of the 35th SIGMOD International Conference on Management

of Data (SIGMOD 2009), Providence, Rhode Island, USA, 29 June – 02 July 2009, pp. 1–2.



16

9. Plattner H., Zeier A. (2011) In-memory data management: An inflection point for enterprise

applications. Springer.

10. Kyriazis D., Varvarigou T., Konstanteli K. (2011) Achieving real-time in distributed computing:

From grids to clouds. IGI Global.

11. Mall R. (2006) Real-time systems: Theory and practice. IGI Global.

12. Laplante P.A., Ovaska S.J. (2011) Real-time systems design and analysis: Tools for the

practitioner. John Wiley & Sons.

13. Ganssle J., Barr M. (2003) Embedded systems dictionary. CMP Books.

14. Oshana R. (2008) Chapter 8: DSP in embedded systems. Embedded software. Elsevier.

15. Gimranov R.D., Agievich V.A. (2013) Providing accurate information in the information

system of a large-scale enterprise using architectural approach. Oil Industry, no 4, pp. 116–119

(in Russian).

16. Gimranov R. (2014) Customer report: Surgutneftegas deploys SAP HANA to increase the

energy efficiency of thousands of operating facilities in real time. SAP Service and Support. SAP

Press, pp. 44–46.

17. Gimranov R.D. (2014) Real-time enterprise 2.0. Changes in corporate information systems

when implementing in-memory data management technology. Mathematics and Information

Technologies in Oil and Gas Industry, no 1, pp. 27–32 (in Russian).

18. Gimranov R.D., Kholkin I.N. (2015) An approach to the controlled evolution of corporate

information systems in the of real time enterprise paradigm. Mathematics and Information

Technologies in Oil and Gas Industry, no 2, pp. 11–28 (in Russian).

19. Gimranov R.D., Kholkin I.N. (2017) Inventing information systems of the future: Theory and

practice. Surgut (in Russian).

20. Williamson O. (1987) The economic institutions of capitalism. Firms, markets, relational

contracting. New York: Free Press.

21. Oleynik A.N. (2000) Institutional economics. Moscow: Infra-M (in Russian).

About the authors

Vladimir I. Ananyin

Senior Lecturer, Department on Business Processes Management,

Russian Presidential Academy of National Economy and Public Administration,

82, Prospect Vernadskogo, Moscow 119571, Russia;


Konstantin V. Zimin

Editor-in-Chief, Information Management Journal;

Member of the Board, The Russian Union of CIO, 34, Seleznevskaya Street, Moscow 123056, Russia;




17

Rinat D. Gimranov

Head of IT Department, PJSC Surgutneftegaz,

1 block 1, Grigoriya Kukuevitskogo Street, Surgut 628415, Russia;


Mikhail I. Lugachev

Dr. Sci. (Econ.), Professor;

Head of Department of Economic Informatics, Lomonosov Moscow State University,

1 build. 46, GSP-1, Leninskie Gory, Moscow 119991, Russia;

Academic Supervisor, IBS Corporate University;


Kirill G. Skripkin

Associate Professor, Department of Economic Informatics, Lomonosov Moscow State University,

1 build. 46, GSP-1, Leninskie Gory, Moscow 119991, Russia;




18

Developing a distributed linguistic decision making system

Alexander V. DemidovskijE-mail: [email protected]

Eduard A. Babkin E-mail: [email protected]

National Research University Higher School of Economics Address: 25/12, Bolshaya Pecherskaya Street, Nizhny Novgorod 603155, Russia

Abstract

In this paper, a new approach to multi-criteria decision making is proposed based on linguistic information taken from a group of autonomous experts. This approach provides an opportunity to better analyze and fi nd solutions for poorly structured problems with consideration of their multidimensionality and uncertainty of context. One of the key components of the proposed methodology is the hierarchy of abstractions proposed by John van Gigch, which presents the levels of alternative solutions and criteria for assessing them. By integrating this hierarchy, it is claimed that the problem situation can be comprehensively analyzed. Therefore, we call our approach multi-level multi-attribute linguistic decision making (ML–MA–LDM).

Our approach includes a methodology that is the particular sequence of steps and the mathematical model, as well as the method to automatically distribute weights of experts’ assessments depending on their confi dence level. Furthermore, this novel approach supports both qualitative and quantitative assessments that are strictly propagated through the complete decision making process across all hierarchical levels of abstraction. Finally, we demonstrate a prototype of a multi-agent expert system for solving poorly structured models with regard to their context uncertainty and multiple aspects. This prototype plays the role of simulation engine for competitive solutions and for verifi cation purposes of the proposed methodology.

Capabilities of the developed approach and the prototype were demonstrated in a practical case of solving a complex confl ict problem of strategic management, as well as rigorous analysis of the proposed approach strengths and weakness that defi nes the direction for further research.

Key words: linguistic decision making; multi-criteria choice; meta-decisions; multi-agent systems;

fuzzy logic; poorly structured problems; decision support systems.

Citation: Demidovskij A.V., Babkin E.A. (2019) Developing a distributed linguistic decision making system.

Business Informatics, vol. 13, no 1, pp. 18–32.

DOI: 10.17323/1998-0663.2019.1.18.32

DATA ANALYSIS AND INTELLIGENCE SYSTEMS


19

Introduction

In the modern world, there is a huge num-

ber of very complex and intricate prob-

lems, such as global warming, hunger,

poverty, unemployment. These problem situa-

tions can be divided into two groups: structured

and poorly structured situations [1]. The latter

are characterized by uncertainty, environmen-

tal variability etc. A large subset of poorly struc-

tured problems can be characterized by a huge

number of stakeholders (or experts), alterna-

tive solutions and criteria, which are used by

decision makers. It is proposed that selection of

one of these alternatives lets a decision maker

solve a problem situation and satisfy a major-

ity of stakeholders. Therefore, creation of new

decision making models and the software design

of expert systems for multi-criteria choice is a

highly topical scientific and social problem.

Moreover, such problem situations frequently

have multiple analysis aspects (or dimensions),

like political (e.g. political tension), econom-

ical (e.g. benefit), ethical (e.g. conformity to

morality) etc. In this way a case of multi-cri-

teria decision making problem appears [2, 3].

The search for the solution of the problem

that has an impact on multiple stakehold-

ers requires mathematical models, algorithms

and a methodology which allow one to analyze

subjective experts’ evaluations from different

aspects. We may note that frequently different

problems’ aspects are hierarchically structured.

In our approach, for multi-criteria choice we

propose to use the framework of meta-deci-

sions suggested by J. van Gigch [4]. We adopt

his main idea of extracting eight abstraction

levels which characterize the principal aspects

of the problematic situation.

There are numerous attempts to elaborate

new decision making approaches or adopt

existing ones to real-life cases, like healthcare

[5], performance evaluation of partnerships

[6], fiber composites optimization [7], reverse

logistics selection and evaluation [8], project

resources scheduling [9], supplier selection

[10], aircraft incident analysis [11]. Usually

traditional approaches like TOPSIS, ELEC-

TRE, VIKOR are used. The considerable

drawback is that these methods rely mostly on

quantitative evaluations, even given in a form

of fuzzy sets [12]. On the other hand, estima-

tions that are given by experts during problem

discussion can be both quantitative and qual-

itative. Qualitative evaluations become more

and more preferable in complex situations

because compared to quantitative evaluations,

qualitative ones have the serious advantage of

their ability to express fuzzy information (e.g.

hesitation). However, according to our rigor-

ous analysis of the field, there is an emerging

trend of combining traditional decision mak-

ing approaches with methods of processing

qualitative evaluations. The combination of

TOPSIS methodology and 2-tuple model for

analyzing qualitative assessments represents a

bright example [13].

Reliable and flexible means for analysis of

qualitative evaluations are provided within the

scientific area of "linguistic decision making"

[2, 3, 14–17] and "linguistic multi-attribute

decision making" [2]. These and other meth-

ods of processing qualitative evaluations now

are generally called "computing with words"

[16–20]. The three most popular approaches

used for calculation in linguistic terms [21]

are:

linguistic computational model based on

membership functions;

linguistic symbolic computational model

based on ordinal scales;

max-min operators, linguistic symbolic

computational model based on convex combi-

nations.

In many cases, information that comes from

the experts is heterogeneous due to its multi-

granularity and there are approaches which

provide methods to work with such informa-

tion: the fusion approach for managing mul-

tigranular linguistic information [22], the lin-



20

guistic hierarchy approach [23] and the method

of extended linguistic hierarchies [14].

This paper presents results of the develop-

ment of a new approach to multi-criteria lin-

guistic decision making in the presence of mul-

tiple hierarchically ordered problem aspects.

Our approach includes a methodology and pro-

totype of a multi-agent expert system for solv-

ing poorly structured models with regards to

their context uncertainty and multiple aspects.

The main contribution in the development of

methods of multi-criteria problem analysis is

development of new scientific principles for

integrating linguistic decision making and the

meta-decision framework of J. van Gigch. This

integration provides stakeholders with a struc-

tured method to analyze the problem from

multiple aspects so that the solution found is

more likely to be objective and optimal than

one that is taken without considering its influ-

ence on all aspects of our life.

This paper has the following structure. In Sec-

tion 1, we provide necessary background infor-

mation that contains a description of basic ele-

ments of the proposed methodology. Then, in

Section 2, we give a detailed description of the

proposed approach which defines the process of

decision making. In Section 3, we demonstrate

the applicability of the proposed approach to the

real case of complex conflict situation in the rice

industry. Section 4 covers details on the design

of a multi-agent system (MAS) that was built for

demonstrating the work of the proposed meth-

odology. Finally, the Conclusion displays the

analysis of the proposed approach and potential

directions of further research.

1. Background and related research

Modeling, analysis and solving poorly struc-

tured problems on the basis of linguistic esti-

mation use several important mathematical

structures.

Definition 1. The linguistic variable is char-

acterized by the tuple:

(H, T (H), U, G, M),

where H – the name of the variable;

T (H) or just T – a set of notions H, i.e. a set

of names of linguistic values H, where each

value is a variable which is denoted in general

case as X and gets values from the set of terms

of the subject area U, which is denoted as u;

G – syntax rule (often takes the form of gram-

mar) for generation of values from H;

M – semantic rule, which defines relation

between H, M (x) [24].

In order to use such linguistic evaluations, it

is important to pick up linguistic descriptors

for a set of concepts and also to define gran-

ularity of uncertainty. Usually the set of con-

cepts is denoted as S = {s0, …, s

g}. The granular-

ity degree of such a set depends on the context

of the problem situation.

On the basis of the given definitions, Herrera

et al. [25] proposed a classical model of analy-

sis of linguistic evaluations using the structure

which is called 2-tuple.

1.1. The classical model on the basis of 2-tuple structure

2-tuple includes the pair [25]:

si S = {s

0, …, s

g} – a linguistic concept;

– a numeric value, or "symbolic trans-

lation", which shows the result of the member

function, i.e. the nearest concept si S = {s

0, …,

sg}, if s

i is not the precise mapping of the given

result.

Later multiple authors proposed a huge num-

ber of operators [3], which allows us to aggre-

gate linguistic information.

1.2. The modernized 2-tuple model

The main problem of the classical model is

the necessity to define the basic scale of evalu-

ations and rules of translation of these evalua-

tions to a single scale. The selection of the scale

and translation rules in that scale becomes



21

a separate and complex task. In their recent

paper [26] researchers proposed a model which

allows one to work with multiple scales with-

out additional transformations. The signifi-

cant difference between the classical model

[25] and the modernized one [26] is the set of

translation rules from the 2-tuple structure to

the numeric representation and vice versa. It is

important to emphasize that this model does

not imply the fact that alternatives and crite-

ria can vary across the time, since it is consid-

ered in a model with bipolar linguistic term

sets [27]. The modernized 2-tuple model [26]

is used in the approach proposed in this paper.

Definition 2. Translation function [26]. Let

S = {s0, …, s

g} be the set of linguistic concepts,

– the set of 2-tuple structures, g = + 1 – its

granularity, – a normalized result of the sym-

bolic aggregation. Then the translation func-

tion can be defined as:

(1)

where round is a function that assigns to the

nearest integer value i {0, 1, ..., g} to .

Definition 3. Reverse translation function

[26]. Let S = {s0, …, s

g} be the set of linguis-

tic concepts, – the set of 2-tuple structures,

g = + 1 – its granularity, (si , – a 2-tuple

structure on , where . Then

the function always exists, so that for the

given 2-tuple structure it returns an equivalent

numeric value [0, 1):

(2)

1.3. 2-tuple model for the comparative

linguistic information

It is reasonable to suppose that experts are

not able to estimate alternatives by a given cri-

teria equally well. When experts are not able to

give precise evaluation, they can make it com-

parative and even express it as a whole sentence

that can have the following structure: "< > is

better than | equal to | worse than < >". This

idea exactly is the basis of the approach which

is called HFLTS (hesitant fuzzy linguistic term

sets) [28].

Definition 4. HFLTS [29]. Let S = {s0, …, s

g}

be a set of linguistic concepts. Then HFLTS or

is an ordered finite set of consecutive linguistic

concepts from S:

HS = {s

i , s

i +1, …, s

j }, S

k S, k {1, ..., g} (3)

In order to avoid information loss when using

HFLTS, another approach was proposed that

is called hesitant 2-tuple set [26]. There are

also operators for aggregation and comparison

of hesitant 2-tuples sets entities: MTWA [26],

MHTWA [26], etc.

Definition 5. Hesitant 2-tuple set [26]. Let

S = {s0, …, s

g} be a set of linguistic concepts,

is a 2-tuple structure on S, i = 1, 2, ..., n. If

(bi ,

i ) < (b

j ,

j )( for any (i < j), (b

1 ,

1 ), (b

2 ,

2 ),

…, (bl ,

l ), which is denoted as T

S , is hesitant

2-tuple set for any i < j . Then HFLTS or HS is

an ordered finite subset of consecutive linguis-

tic concepts from S.

1.4. A meta-decision framework

for analysis of problem situations

from different abstraction levels

Due to the fact that during the process of

finding solutions for complex problems it is

important to analyze the situation from differ-

ent aspects, we decided to use eight abstrac-

tion levels that were initially proposed by

J. van Gigch in his meta-decision framework

[26]. These levels are used as the basic set of

aspects of any analyzed problem. More spe-

cifically, these levels are (in increasing order of

abstraction level): managerial, economic, sci-

entific, legal, political, epistemological, ethi-

cal, aesthetic.



22

Definition 6. Abstraction is a mental process

in which representations of reality are defined

on different levels of conceptualization.

Definition 7. An abstraction level (a logic

level) – a perspective or a point of view from

which stakeholders are trying to solve the prob-

lem. A chosen perspective reflects historical

skills of an expert on the given abstraction level

(the logic level).

2. Proposed multi-criteria decision making approach

In the previous chapter, basic linguistic deci-

sion making (LDM) approaches were described

as well as eight levels of abstraction that are vital

for analysis of complex problems. It is impor-

tant to emphasize that existing approaches con-

centrate either on analysis of only quantita-

tive assessments or only qualitative ones. Very

few approaches focus on both types of estima-

tions. At the same time, modern methodologies

are likely to assume that there are a number of

experts without capturing the area of their exper-

tise as well as the fact that criteria also belong to

different abstraction levels, like politics, econom-

ics etc. More importantly, existing methods for

decision making are demonstrated on artificial

cases with very few experts and alternative solu-

tions. Finally, the demonstration is never made

in the dynamics of a multi-agent system (MAS),

although not only could it help to reveal draw-

backs of existing approaches but also to analyze

the behavior of agents and details of their interac-

tion. For example, it is promising to also consider

trust among experts. This brings us to the point to

propose a new methodology which could incor-

porate most of the gaps described above.

In this section, we will describe the proposed

approach for solving poorly structured prob-

lems that are capable of taking into considera-

tion multiple hierarchically ordered aspects of

the problem situation and process heterogene-

ous evaluations. We call our approach multi-

level multi-attribute linguistic decision making

(ML–MA–LDM).

2.1. Description of steps during ML–MA–LDM

The proposed approach consists of several

consecutive steps starting from defining the

estimation rules and finishing with the com-

munication stage (Figure 1). It is important to

note that these steps can be found individually

in various papers describing the decision mak-

ing process, for example in [30, 31], but never

were fused in a consistent way. The proposed

approach includes:

1. Setting up rules for providing estimations

and distribution of criteria weights. In the pro-

posed approach we make several assumptions:

а. experts give honest evaluations;

в. experts believe each other;

Definition of estimation rules

Formulating desired states

Formulating criteria

Formulating alternative solutions

Multi-Level Multi-Attribute estimating

Aggregation of estimations

Search for the best alternative solution

Communication of a solution found

Defin

ition

of i

nitia

l lin

guis

tic d

ata

Fig. 1. The proposed methodology to solve poorly structured problems in conditions

of uncertainty of context and fuzzy estimations



23

с. experts choose granularity of evaluations

according to their experience and knowl-

edge about a problem;

d. experts have the same understanding of

evaluations;

2. Defining available linguistic sets, a context-

free grammar and transformation function;

3. Multi-level definition of the desired state,

criteria and alternatives.

a. analyzing the desired state on each level

of abstraction;

b. formulating criteria for each level of

abstraction;

c. formulating alternatives.

4. Giving multi-level and multi-criteria evalu-

ations.

a. aggregating information;

b. searching for the best alternative;

c. communicating the solution found.

2.2. Aggregating information

After criteria and alternatives were defined,

all experts start giving evaluations of each alter-

native for each available criterion.

Let x = {x1, x

2, ..., x

N } is the list of alternatives,

c = {c1, c

2, ..., c

M } is the list of criteria, e = {e

1,

e2, ..., e

T } is the list of experts. We assume that

each expert ek can evaluate alternatives using

different linguistic scales Sg k with granular-

ity gk. In the case of comparative evaluations,

we also have the grammar GH which can be

also used for creation of linguistic evaluations.

Moreover, the criteria are given for each level

of abstraction in the meta-decision framework,

i.e. let l = {l1, l

2, ..., l

Z } be the list of the levels of

abstraction.

The overall sequence of steps is described in

Figure 2. These steps describe pre-processing

and aggregation of evaluations collected from

experts. Therefore, as a result, one evaluation

for each given alternative is obtained and the

best alternative can be found by sorting these

evaluations according to rules of comparing

hesitant 2-tuple fuzzy sets.

Step 1. Formulating matrices of HFLTS

evaluations. Due to the fact that experts can

give evaluations in a different form, it is impor-

tant to preprocess them. More specifically,

evaluations should be translated to HFLTS as

this format is flexible enough to represent both

precise and interval evaluations. As a result, for

each expert we get a matrix of evaluations

,

where – an evaluation of the expert ek for

the i-th alternative on the j-th criterion in the

format of HFLTS on the scale Sg.

Step 2. Aggregation of evaluations by cri-

teria. During this step, it is important to find

an accumulated evaluation for combination of

each alternative i, every level of abstraction l,

and every expert ek by aggregating evaluations

for every criterion corresponding to the given

abstraction level. Then for each expert we get a

following matrix:

, (4)

where i – the index of alternative;

j – the index of the abstraction level;

Translating estimations to hesitant 2-tuple sets

Aggregating estimations on the criteria level

Translating estimations to abstractions level

Aggregating estimations on the experts level

Aggregating estimations on the abstractions level

Fig. 2. A structure of the “Aggregating information” step of the proposed methodology



24

p – the vector of criteria weights,

.

Here we propose to use the MHTMA opera-

tor because each criterion has its own defined

weight. So, for each expert we get the following

decisions matrix:

,

where – the evaluation of the expert ek for

i-th alternative for j-th level of abstraction in a

form of HFLTS on the scale Sg k.

Step 3. Translation of evaluations to

abstraction levels. The next step should be

aggregation of evaluations for each level of

abstraction separately. From the previous step

we get T matrices with evaluations, each of size

N Z. In order to make aggregation for each

level of abstraction, we need to have Z matrices

with evaluations, each of the size N T, where

N is a number of alternatives and T is a number

of criteria. So, for each abstraction level we get

the following decisions matrix:

,

where – the evaluation for lu-th abstraction

level from the i-th alternative for j-th expert in

a form of HFLTS on the scale Sg k.

Step 4. Aggregation of evaluations by

expert. During this step, the total evaluation is

calculated for each level of abstraction lu, for

each i-th alternative, and for each expert given.

If w is the given vector of experts’ weights,

,

then for each level of abstraction we get the fol-

lowing matrix:

(5)

where i – the index of the alternative;

j – the index of the abstraction level.

If the vector of weights is not given, the fol-

lowing formula should be used for their calcu-

lation:

(6)

where w [0, 1) – the proportion of the first

expert’s evaluation in the weights sum.

Therefore, we get the following decisions

matrix

,

where is aggregated evaluation for i-th

alternative and for j-th level of abstraction in a

form of HFLTS on the scale Sg k.

Step 5. Aggregation of evaluation by levels

of abstraction. During this step the total evalu-

ation for each i-th alternative and for each level

of abstraction is found:

, (7)

where i – the index of alternative; q – the vector of weights of levels of abstraction,

.

So, we get the following vector of evaluations

,

where is the aggregated evaluation for i-th

alternative in a form of HFLTS on the scale Sg k.

As a result, we get assessments that draw

insights on how each alternative is measured on

each level of abstraction and a decision maker

can use this information to better understand

the scope of alternatives and their influence on

each aspect of the problem situation. It can also

be possible to customize a methodology at this

point; for example it is possible to select only

a subset of levels of abstraction which interest

the decision maker to make the final decision.



25

3. Demonstration in one case

For demonstrating our approach, we use a

complex problem situation with rice produc-

tion in the state Chhattisgarh (India) [32]. Rice

is one of the main products in India in terms

of consumption. This state is the biggest pro-

vider of paddies. The first step is to give a gen-

eral description of the current situation.

3.1. Description of the current state

In the Chhattisgarh state, the rice industry

obeys the Government. There is a huge number

of farmers, the majority of whom are middle-

and small-sized households. Middle- and small-

sized households are very dependent on weather

conditions and Government politics with respect

to buying the rice left over at the end of the sea-

son for distribution among poor people. That

is why they have to take loans that often bank-

rupt households. This in turn makes the number

of working population in rice industry decline.

After the rice is ready, farmers sell rice to millers.

Millers do not rush to buy rice since the Govern-

ment buys rice at very low prices at the end of the

season. Millers clean the paddy up, produce rice

and sell it via sales agents. The miller business has

minimal profitability, and that is why the market

is decreasing and only big players are left there.

These big players define the rice price to make it

as low as possible. Rice cannot be exported due

to the use of several fertilizers that damage the

atmosphere. The overall political atmosphere is

unfavorable.

3.2. Description of a desired state

Households receive subsidies from the Gov-

ernment on their business. Rice that is left

unbought at the end of the season is bought at

the market price by either the Government or

millers. The Government prevents the crea-

tion of miller monopolies that tend to reduce

the market price. Moreover, there is an active

export policy that let millers increase their

profits. Moreover, innovative technologies

make it possible to avoid use of polluting fer-

tilizers, thus opening a door for export. Millers

have a joint logistics union that lets them con-

trol the supply chain. The poor get rice from

the Government and this, in turn, motivates

them to become farmers. Low unemployment

decreases chaos on the streets.

Due to the multidimensionality of the prob-

lematic situation, there are a large number

of alternative solutions. Alternative solutions

define the set of actions that can be later evalu-

ated by criteria defined earlier. In order to for-

mulate them there is a specific technique:

1. Definition of the desired state of industry

for each level of abstraction;

2. Definition of criteria specific for each level;

3. Definition of concrete alternative solu-

tions driven by the desired state on each level.

In the given case there are the following

experts: the representative of the Department

of Foreign and Domestic Policies (DFD), the

representative of the Department of social pol-

itics (DSP), the representative of farmers (F),

the owner of a mill (M), a sales agent (SA), a

rice transporter (RT), an ecologist (E).

We consider the experts having experience

on the following levels of abstraction (Table 1):

managerial (MLA), economic (ELA), scien-

tific (SLA), legal (LLA), political (PLA), epis-

temological (EPLA), ethical (ETLA), aes-

thetic (ALA).

3.3. Aggregating information

According to our approach, the follow-

ing actions should be taken for the reasonable

choice of the problem solution.

Step 1. Formulating matrices of evalua-

tions. As HFLTS allows to use multiple lin-

guistic scales and there is no need to translate

evaluations to a single scale, the only needed

transformation is to translate all evaluations

to the form of HFLTS. Let is suppose, that



26

an expert gave the evaluation ("good", "vary

good"). The evaluation can be translated to the

instance of Hesitant 2-tuple Set: ,

where S 7 = – very bad, – bad, – slightly

fair, – fair, – slightly good, –good, –

very good .

After that, all evaluations are in a united form

and it is possible to start aggregating them. It is

mandatory to define weights for criteria and the

levels of abstraction. In this case, because there

are no presuppositions on importance neither

for criteria nor for alternatives, weights are equal

among both the alternatives and the criteria.

Step 2. Aggregating evaluations for criteria.

The very first step is to find the aggregated esti-

mation for every expert, every alternative and

every level of abstraction. Aggregation hap-

pens across criteria which belong to the same

level of abstraction. In our example we assume,

that the expert of Department of Foreign and

Domestic Politics (DFD) gave following esti-

mations for the alternative A.ETLA.1 (Table 2)

on a political level of abstraction (PLA).

For example, we consider the weights of the

criteria to be equal: w = (0.33, 0.33, 0.33).

For calculating an aggregated evaluation, the

MHTWA operator is used:

aggreagated_value =

Step 3. Translation to the levels of abstrac-

tion. This is the technical transformation of

given matrices and it is described in Step 3 of

the proposed methodology.

Step 4. Aggregation of evaluations by

experts. During this step, the accumulated

evaluation for each alternative, each level of

abstraction and each expert is calculated. In

this case, experts’ weights are distributed in a

way that the expert who gives the most precise

evaluation has the bigger weight.

Table 1. Experience of experts participating in the evaluation

MLA ELA SLA LLA PLA EPLA ETLA ALA

DFD x x x x x

DSP x x x

F x x x

M x x x

SA x x x

RT x x

E x x x x

Table 2. DFD evaluations for the alternative A.ETLA.1

Criteria on PLA

C.PLA.1 C.PLA.2 C.PLA.3

A.ETLA.1



27

Step 5. Aggregation of evaluations by lev-

els of abstraction. During this step, evaluations

are accumulated by each level of abstraction

to get the final evaluation for each alternative.

Table 3 shows the results of aggregation for the

described case.

Table 3. The ordered list of alternatives

and accumulated evaluations

Alternative name Estimation

A.ELA.7 Increase crop via irrigation system implementation

A.SLA.3 Decrease usage of fertilizers

A.ELA.2 Increase taxes for farmers

A.ELA.1 Increase subsidies for farmers

Step 6. Seeking the best alternative. Dur-

ing this step, the best alternative is chosen. For

that, the list of calculated evaluations should

be ordered according to the rules of compar-

ing instances of Hesitant 2-tuple Set. In the

described case, the best alternative is the one

with id A.ELA.7 "Increase crop via irrigation

system implementation".

Step 7. Communication of the solution

found. All the participants of the decision

making problem are notified about the solu-

tion found. It is important to draw attention to

the fact that to find the solution, multiple alter-

native solutions were assessed against multiple

criteria and, which is more important, each

alternative solution was analyzed separately

on different level of abstraction representing a

vital aspect of the problem situation.

4. Implementation details

4.1. MAS design and implementation

For validation of the proposed LDM multi-

level model and our approach in general, an

expert system was developed and tested for a

relevant use case. The system was originally

designed as a distributed multi-agent system

(MAS) with a belief-desire-intention (BDI)

architecture [33]. It is a promising set of prin-

ciples for designing an MAS and has practical

use in various projects, like supply chain mod-

eling [34], transport logistics [35] and time-

tabling [36]. During design and implemen-

tation, we exploited advanced features of the

MAS platform JASON1 and its extension JaC-

aMo framework2. JASON provides a power-

ful AgentSpeak interpreter and basic commu-

nication primitives, while JaCaMo offers such

environment artifacts as tasks, bids, etc. New

numerical and linguistic algorithms related to

our proposed LDM multi-level models were

implemented in Java and then were encapsu-

lated to the JASON coordinator agent using

the Java-AgentSpeak proxy. The architecture

of the MAS is presented in Figure 3. A detailed

explanation of the level of implementation is

given in Figure 4.

There are always two types of agents availa-

ble in the system: a coordinator and an expert.

While it is enough to have a single coordina-

tor to rule the whole decision process, there

are multiple expert entities that make evalua-

tions based on the problem context. A number

of experts in simulation represents one-to-one

mapping to experts in the real life.

The coordinator is an agent that has two main

goals: starting the decision making process and

accumulation and calculation of the best alter-

native solution based on the evaluations pro-

vided. At the same time, coordinator activates

the main goal of the expert by publishing the

task in the Common Environment Artifact:

giving evaluations for the given problem on

the basis of alternatives and criteria provided 1 http://jason.sourceforge.net/wp/1 http://jacamo.sourceforge.net/



28

Fig. 4. Jason implementation of MAS

Fig. 3. Multi-agent architecture for multi-attribute LDM

Subscribe

Subscribe Publisch

Legend

Find bestalternative

Bid

Task

Artifact

Bids

Estimatcs

Ecolog Farmer Transporter Miller Miller Agent Politican Social Politican Coordinator

Coordinator

Scales

Expert-specific component

Problem-specific component

Problem-agnosticcomponent

CriteriaProblem

Description

coordinator.asl

2 5

3

4

1

Winneralternative

expert.asl

Achieve:focus (Common Environment) Winner

Expert

Legend

Common Environment

start (task_name, criteria, alternatives)

decide (task_name, criteria, alternatives)

sove_tasts (name, criteria, alternatives)

Bid

Task

Goal Publisch Communicate Belog Agent Artifact

Alternatives



29

by coordinator. Once all the needed evalua-

tions are made, the coordinator tries to achieve

his second goal – finding out what alterna-

tive is best according to our LDM multi-level

model. As for an expert agent, its only goal is to

give evaluations by publishing in the Common

Environment artifact. Both coordinator and

expert agents are subscribed to the entity of the

winner in the Common Environment artifact

and get notified when it appears after all calcu-

lations are done.

4.2. Description of decision making in the MAS expert system

The algorithm of the decision making in our

multi-agent expert system follows the formal

methodology of our approach. During an ini-

tialization phase, the coordinator provides

experts with information on the common envi-

ronment (CE) where they will work together.

Experts also get prepared by subscribing to the

task to be notified when it is published. When

experts get notification about the new task,

they start providing their evaluations of the

given problem situation. Moreover, experts

subscribe to the winner alternative (WA) to

be aware of the best alternative. It is chosen

based on the evaluations of all agents. When

all preparations are done and experts are wait-

ing for the task to appear, the coordinator

publishes the task. All tasks contains the prob-

lem description, alternatives and criteria –

all necessary information for experts to ana-

lyze the problem and evaluate every alterna-

tive by given criteria.

After experts evaluate every alternative solu-

tion of the given problem, they publish bids

that contain these evaluations alongside the

description of scales that were used during

the decision process. These bids are handled

and stored in the common environment. The

coordinator either waits for all experts to pro-

vide evaluations or waits for a certain, explic-

itly defined period and then closes the admis-

sion. As soon as the admission is closed, the

coordinator initiates accumulation of all the

evaluations that is performed according to

the formal algorithm proposed in this paper.

When the calculations are finished, the win-

ning alternative (WA) is published and every

expert is notified about it. This appears to be

the end of the simulation, however the sys-

tem can be still active and waiting for a new

request.

The implementation of algorithms of aggre-

gation of heterogeneous estimations was

aligned with corporate enterprise standards

of software development. Furthermore, the

authors elaborated the input/output format

for describing the important parameters (crite-

ria, alternatives, levels, experts). The software

implementation of the prototypes is available

publicly on GitHub3 and contains the com-

plete system described in Figure 3. It can be

further extended for a more general case.

Conclusion

In the framework of current research, we

have made a broad investigation of the field

and aligned research with design science [37]

methodology. Rigorous analysis of existing

approaches to linguistic multi-criteria deci-

sion making revealed their disunity and inferi-

ority if applied to problems with heterogene-

ous information and uncertainty of context.

On the one hand, there are classical decision

making approaches that instruct each expert

to find the best alternative, however quanti-

tative estimations are not taken into consid-

eration. On the other hand, methods of LDM

are supposed to tackle heterogeneous estima-

tions, though they are hardly applied to real

life problems due to lack of unified method-

ology for searching for the best alternative.

More importantly, poorly structured prob-

lems are characterized by a huge number of

stakeholders. 3 https://github.com/demid5111/lingvo-dss-bdi



30

References

1 . Zopounidis C., Doumpos M. (2000) Intelligent decision aiding systems based on multiple criteria for financial engineering. Springer Science & Business Media.

2. Xu Z. (2012) Linguistic decision making. Theory and methods. Springer-Verlag Berlin Heidelberg.

3. Martinez L., Rodriguez R., Herrera F. (2015) The 2-tuple linguistic model: Computing with words in decision making. Cham: Springer.

4. Van Gigch J. (2003) Metadecisions: Rehabilitating epistemology. Springer Science & Business Media.

5. Dehe B., Bamford D. (2015) Development, test and comparison of two multiple criteria decision

analysis (MCDA) models: A case of healthcare infrastructure location. Expert Systems with Applications,

vol. 42, no 19, pp. 6717–6727.

In our approach, we propose to extend tra-

ditional LDM methods with the meta-deci-

sion framework on the basis of the abstraction

hierarchy by J. van Gigch that suggests analyz-

ing the problem from eight different perspec-

tives.

The proposed approach has a set of improve-

ments compared to existing approaches in

LDM. As was already stated, the considerable

drawback of existing approaches is that they

concentrate either on analysis of only quanti-

tative assessments, e.g. TOPSIS, ELECTRE,

VIKOR etc., or only qualitative ones [25, 26,

29]. Very few approaches focus on both types

of evaluations [13] or fuzzy sets [12]. Our pro-

posed approach relies on both types of assess-

ments originally from the very beginning. At

the same time, modern methodologies do

not respect the fact that experts differ in their

knowledge in different areas, like politics,

economics etc. Moreover, our approach offers

a reliable mechanism for selecting weights for

experts’ evaluations depending on how pre-

cise their evaluations are. We also proposed a

format for describing poorly structured prob-

lems. Existing approaches are demonstrated

on artificial cases that are far from reality.

The proposed approach was demonstrated

on a real life case that has an impact on huge

number of stakeholders, while existing meth-

ods are demonstrated on artificial cases with

very few experts and alternative solutions. In

addition to that, the implementation of the

proposed approach was performed as well as

implementation of algorithms for aggregating

heterogeneous information where there is no

existing open source implementation. Finally,

the implementation is made in dynamics of

a multi-agent system (MAS). This allows a

decision maker to analyze behavior of agents

and details of their interaction. For example,

it is promising to also consider trust among

experts. This brings us to the point to pro-

pose a new approach which could bridge

most of the gaps described above. We could

not find any similar solution among existing

approaches and consider it to be an impor-

tant innovation in the field of decision mak-

ing, because it can be used in the context of

various poorly structured tasks with multiple

stakeholders and alternatives – all of which is

especially topical in the modern world. Our

expert system demonstrated the desired out-

come for the case of a complex problem situa-

tion with rice production.

An important step of any research is the def-

inition of the further direction of study. The

authors suggest the following improvement of

their ideas to get rid of the drawbacks of the

proposed approach:

1. Consider the trust factor among experts;

2. Consider ontologies of difference among

experts. These differences are expressed in var-

ious meanings of the same linguistic concepts

for multiple experts. In the proposed method-

ology, authors make an assumption that ontol-

ogy is the same for all experts.



31

6. Piltan M., Sowlati T. (2016) A multi-criteria decision support model for evaluating the performance of

partnerships. Expert Systems with Applications, vol. 45, pp. 373–384.

7. Alemi-Ardakani M., Milani A.S., Yannacopoulos S., Shokouhi G. (2016) On the effect of subjective,

objective and combinative weighting in multiple criteria decision making: A case study on impact

optimization of composites. Expert Systems with Applications, vol. 46, pp. 426–438.

8. Senthil S., Srirangacharyulu B., Ramesh A. (2014) A robust hybrid multi-criteria decision making

methodology for contractor evaluation and selection in third-party reverse logistics. Expert Systems with Applications, vol. 41, no 1, pp. 50–58.

9. Markou C., Koulinas G.K., Vavatsikos A.P. (2017) Project resources scheduling and leveling using

multi-attribute decision models: Models implementation and case study. Expert Systems with Applications, vol. 77, pp. 160–169.

10. Cid-Lopez A., Hornos M.J., Carrasco R.A., Herrera-Viedma E. (2016) Applying a linguistic

multi-criteria decision-making model to the analysis of ICT suppliers’ offers. Expert Systems with Applications, vol. 57, pp. 127–138.

11. Skorupski J. (2014) Multi-criteria group decision making under uncertainty with application to air traffic

safety. Expert Systems with Applications, vol. 41, no 16, pp. 7406–7414.

12. Igoulalene I., Benyoucef L., Tiwari M.K. (2015) Novel fuzzy hybrid multi-criteria group decision

making approaches for the strategic supplier selection problem. Expert Systems with Applications, vol. 42,

no 7, pp. 3342–3356.

13. Cid-Lopez A., Hornos M.J., Carrasco R.A., Herrera-Viedma E., Chiclana F. (2017) Linguistic

multi-criteria decision-making model with output variable expressive richness. Expert Systems with Applications, vol. 83, pp. 350–362.

14. Espinilla M., Liu J., Martinez L. (2011) An extended hierarchical linguistic model for decision-making

problems. Computational Intelligence, vol. 27, no 3, pp. 489–512.

15. Herrera F., Herrera-Viedma E. (2000) Linguistic decision analysis: steps for solving decision problems

under linguistic information. Fuzzy Sets and Systems, vol. 115, no 1, pp. 67–82.

16. Martinez L., Ruan D., Herrera F., Herrera-Viedma E., Wang P.P. (2009) Linguistic decision making:

Tools and applications. Information Sciences, vol. 179, no 14, pp. 2297–2298.

17. Yager R.R. (1981) Concepts, theory, and techniques a new methodology for ordinal multiobjective

decisions based on fuzzy sets. Decision Sciences, vol. 12, no 4, pp. 589–600.

18. Martinez L., Da R., Herrera F. (2010) Computing with words in decision support systems: An overview

on models and applications. International Journal of Computational Intelligence Systems, vol. 3, no 4,

pp. 382–395.

19. Martinez L., Herrera F. (2012) An overview on the 2-tuple linguistic model for computing with words in

decision making: Extensions, applications and challenges. Information Sciences, no 207, pp. 1–18.

20. Yager R.R. (1995) An approach to ordinal decision making. International Journal of Approximate Reasoning, vol. 12, no 3–4, pp. 237–261.

21. Xu Z. (2004) A method based on linguistic aggregation operators for group decision making with

linguistic preference relations. Information sciences, vol. 166, no 1–4, pp. 19–30.

22. Herrera F., Herrera-Viedma E., Martinez L. (2000) A fusion approach for managing multi-granularity

linguistic term sets in decision making. Fuzzy Sets and Systems, vol. 114, no 1, pp. 43–58.

23. Herrera F., Martinez L. (2001) A model based on linguistic 2-tuples for dealing with multigranular

hierarchical linguistic contexts in multi-expert decision-making. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), vol. 31, no 2, pp. 227–234.

24. Zadeh L. (1979) A theory of approximate reasoning. Machine Intelligence, no 9, pp. 149–194.

25. Herrera F., Martinez L. (2000) A 2-tuple fuzzy linguistic representation model for computing with words.

IEEE Transactions on Fuzzy Systems, vol. 8, no 6, pp. 746–752.



32

26. Wei C., Huchang L. (2016) A multigranularity linguistic group decision-making method based on

hesitant 2-tuple sets. International Journal of Intelligent Systems, vol. 31, no 6, pp. 612–634.

27. Liu H., Le J., Martinez L. (2018) A dynamic multi-criteria decision making model with bipolar linguistic

term sets. Expert Systems with Applications, vol. 95, pp. 104–112.

28. Torra V. (2010) Hesitant fuzzy sets. International Journal of Intelligent Systems, vol. 25, no 6,

pp. 529–539.

29. Rodriguez R., Martinez L., Herrera F. (2012) Hesitant fuzzy linguistic term sets for decision making.

IEEE Transactions on Fuzzy Systems, vol. 20, no 1, pp. 109–119.

30. Mardani A., Jusoh A., Zavadkas E.K. (2015) Fuzzy multiple criteria decision-making techniques and

applications – Two decades review from 1994 to 2014. Expert Systems with Applications, vol. 42, no 8,

pp. 4126–4148.

31. Nguyen H.-T., Dawal S.Z.M., Nukman Y., Aoyama H. (2014) A hybrid approach for fuzzy multi-

attribute decision making in machine tool selection with consideration of the interactions of attributes.

Expert Systems with Applications, vol. 41, no 6, pp. 3078–3090.

32. Garg M. (2012) Cases on supply chain and distribution management: Issues and principles. IGI Global.

33. Bratman M. (1987) Intention, plans, and practical reason. Cambridge, MA: Harvard University Press.

34. Baykasoglu A., Kaplanoglu V. (2015) An application oriented multi-agent based approach to dynamic

load/truck planning. Expert Systems with Applications, vol. 42, no 15, pp. 6008–6025.

35. Satunin S., Babkin E. (2014) A multi-agent approach to intelligent transportation systems modeling with

combinatorial auctions. Expert Systems with Applications, vol. 41, no 15, pp. 6622–6633.

36. Babkin E., Abdulrab H., Babkina T. (2009) AgentTime: A distributed multi-agent software system for

university’s timetabling. From System Complexity to Emergent Properties. Berlin: Springer, pp. 141–154.

37. Peffers K., Tuunanen T., Rothenberger M.A., Chatterjee S. (2007) A design science research

methodology for information systems research. Journal of Management Information Systems, vol. 24,

no 3, pp. 45–77.

About the authors

Alexander V. Demidovskij

Doctoral Student, Department of Information Systems and Technologies,

National Research University Higher School of Economics,

25/12, Bolshaya Pecherskaya Street, Nizhny Novgorod 603155, Russia;


Eduard A. Babkin

Cand. Sci. (Tech.), PhD (Computer Science);

Professor, Department of Information Systems and Technologies,

National Research University Higher School of Economics,

25/12, Bolshaya Pecherskaya Street, Nizhny Novgorod 603155, Russia;




33

Developing parallel real-coded genetic algorithms for decision-making systems of socio-ecological and economic planning

Andranik S. Akopov a,b


Armen L. Beklaryan a,b


Manoj Thakur c


Bhisham Dev Verma c


a National Research University Higher School of Economics Address: 20, Myasnitskaya Street, Moscow 101000, Russia b Central Economics and Mathematics Institute, Russian Academy of Sciences Address: 47, Nakhimovky Prospect, Moscow 117418, Russia c Indian Institute of Technology Mandi Address: Mandi, Himachal Pradesh 175005, India

Abstract

This article presents a new approach to designing decision-making systems for socio-economic and ecological planning using parallel real-coded genetic algorithms (RCGAs), aggregated with simulation models by objective functions. A feature of this approach is the use of special agent-processes, which are autonomous genetic algorithms (GAs) acting synchronously in parallel streams and exchanging periodically by the best potential decisions. This allows us to overcome the premature convergence problem in local extremums. In addition, it was shown that the combined use of diff erent crossover and mutation operators signifi cantly improves the time effi ciency of RCGAs, as well as the quality of the decisions obtained (proximity to optimum), providing a more diverse population of potential decisions (individuals).

In this paper, several suggested crossover and mutation operators are used, in particular, a modified simulated binary crossover (MSBX) and scalable uniform mutation

MODELING OF SOCIAL AND ECONOMIC SYSTEMS


34

Introduction

Currently, there is a need to design

decision-making systems for socio-

economic and environmental plan-

ning using simulation models aggregated with

genetic optimization algorithms for solving

large-scale optimization problems.

For the first time, a similar approach was pro-

posed in [1–3] in which the developed multi-

agent genetic optimization algorithm MAG-

AMO was presented. MAGAMO is aggregated

through objective functions with a simulation

model of a distance trading enterprise. In [1],

operator (SUM), which is based on quantization of the feasible region of the search space (dividing the feasible region on small subranges with equal lengths) while taking into account the common amount of interacting agent-processes and the maximum number of internal iterations of GAs forming potential decisions through selection, crossover and mutation. Such a functional dependence of the parameters of heuristic operators on the corresponding process characteristics, aggregated with the combined probabilistic use of various crossover and mutation operators, makes it possible to get maximum effect from the multi-processes architecture. As a result, the computational possibilities of RCGAs for solving large-scale optimization problems (hundreds and thousands of decision variables, multiple objective functions) become dependent only on the physical characteristics of the existing computing clusters. This makes it possible to efficiently use supercomputer technologies.

An important advantage of the proposed system is the implemented integration between the developed parallel RCGA (implemented in C++ and MPI) and the simulation modelling system AnyLogic (Java) using JNI technology. Such an approach allows one to synthesize real world optimization problems in decision-making systems of socio-economic and ecological planning, using simulation methods supported by AnyLogic. The result is an eff ective solution to single-objective and multi-objective optimization tasks of large dimension, in which the objective functionals are the result of simulation modeling and cannot be obtained analytically.

Key words: real-coded genetic algorithms; multi-objective optimization; Pareto front; simulation

modeling; AnyLogic.

Citation: Akopov A.S., Beklaryan A.L., Thakur M., Verma B.D. (2019) Developing parallel

real-coded genetic algorithms for decision-making systems of socio-ecological and economic

planning. Business Informatics, vol. 13, no 1, pp. 33–44

DOI: 10.17323/1998-0663.2019.1.33.44

the objective functions, in particular, were the

accumulated profit, the size of the active client

base and the inventory turnover. At the same

time, a similar model included five product cat-

egories, six cities and three customer segments,

which, taking into account multiple restric-

tions and temporal granularity, characterized it

as a large-scale optimization problem.

Note that MAGAMO [3] uses the dynamic

interaction of synchronized intelligent agents,

each of which is an autonomous genetic algo-

rithm (GA) that implements an internal pro-

cedure for the formation of an archive of



35

non-Pareto solutions, for example, SPEA2

(Strength Pareto Evolutionary Algorithm)

[4]. In MAGAMO, the dimensionality of the

solved optimization problem is reduced by

splitting the initial set of decision variables

into small groups with their subsequent distri-

bution between process agents (autonomous

GAs) to minimize the size of local populations

and the number of necessary (resource-inten-

sive) recalculations of fitness function values,

respectively. The MAGAMO algorithm was

previously used, in particular, for the rational

control of environmental modernization of

enterprises that are stationary sources of harm-

ful emissions [5], to determine the best geolog-

ical and technical activities on wells [6], etc. In

[7], some modification of this heuristic algo-

rithm through inclusion of an adaptive mecha-

nism provided improving values of GA param-

eters on the individual level of agent-processes

depending on the optimization results (values

of minimized target of functions, the rate of

convergence, the hypervolume metric of the

Pareto front, etc.).

At the same time, a significant deficiency

of MAGAMO is the use of binary coding of

decision variables values, which causes use of

classical operators of a single-point and two-

point crossover, as well as inversion (binary)

mutation. As a result, the time-efficiency of

the algorithm reduction if there is a need to

search for solutions in a continuous space of

high dimensionality, i.e. when wide values of

feasible ranges are specified for decision vari-

ables (for example, [–100, 100]), and there

are increased requirements for the precision of

computations (when the number of bits of the

mantissa is 2 or more).

Another problem is the weak mutual aggre-

gation of agent-processes in MAGAMO and

the need to synchronize their states (replica-

tion of the values of decision variables between

processes) at each GA iteration, all of which

significantly reduces the efficiency of process

parallelization.

Therefore, it is necessary to create a funda-

mentally new parallel genetic algorithm using

the mechanism of real coding, i.e. belonging

to the class of RCGA algorithms (real-coded

genetic algorithms) [8] and this is based on

using new heuristic operators of the appro-

priate type providing a mechanism of peri-

odic exchanges of the best potential decisions

between agents-processes.

The purpose of this paper is to develop a

multi-agent parallel real-coded genetic algo-

rithm for solving multi-objective optimiza-

tion problems (MA–RCGA–MO) aggregated

through the objective functions with AnyLogic

simulation models. In the result, there is provi-

sion for solving large-scale optimization prob-

lems in decision-making systems for socio-

economic and environmental planning.

It should be noted that the choice of the Any-

Logic system is mainly prompted by the impor-

tant advantages of the platform, such as sup-

porting system dynamics methods, discrete

event modeling and agent-based modeling

within one model [9]. This allows you to design

decision-making systems that require the devel-

opment and use of complex simulation models.

Examples are the system of rational control of

ecological-economic systems [10–13], the sys-

tem of modelling and optimization of the set of

investments of a vertically integrated oil com-

pany [14–15], the system of optimal distribu-

tion of flows of requests for loans at the inter-

regional underwriting center of a very large

bank [16], the system of control of intellectual

agent-rescuers behavior in the simulation of

human crowd behavior in an emergency [17–

18] and other systems.

1. Multi-agent parallel real coded genetic algorithm

Currently there is a line of well-known

research on genetic optimization algorithms

designed to solve multi-objective optimiza-

tion problems. Among the most often used

methods, the following algorithms should be



36

highlighted: SPEA2 [4], MOEA/D (multi-

purpose GA based on decomposition) [19],

NSGA-II and NSGA-III (multi-objective GA

based on non-dominated sorting) [20, 21] and

some other algorithms. In addition, the appli-

cation of agent-based modelling for imple-

menting GA is known as well. In particular,

the MAGA algorithm [22] that is intended

for solving large-scale optimization problems

should be noted. Despite the multiple advan-

tages of developed GAs, most of them use the

binary coding mechanism for decision varia-

bles, which causes significant loss of time-effi-

ciency when searching for solutions in a large-

scale continuous space. Accordingly, this limits

the possibility of using such GAs in designing

decision-making systems based on simulation

modelling of the behavior of complex objects.

In order to overcome these difficulties, a new

multi-agent parallel real-coded genetic algo-

rithm is proposed. The algorithm is intended

for solving large-scale multi-objective optimi-

zation problems.

The main features of the suggested algorithm

are the following:

using well-known crossover and mutation

operators designed for real-coded genetic algo-

rithms (RCGAs), such as SBX crossover (sim-

ulated binary crossover) [23], Laplace crosso-

ver (LX) [24], power mutations (PM) [25] and

others;

using new (modified) heuristic crosso-

ver and mutation operators, the characteris-

tics of which functionally depend on the indi-

vidual number of the associated agent-process

(i.e., the process in which they are performed).

This makes it possible to significantly improve

their efficiency, in particular, to achieve bet-

ter diversity of potential decisions, to provide

splitting (quantization) of search ranges into a

larger number of short intervals and, thus, to

use maximally the capabilities of a multi-clus-

ter (multiprocessor) computing system while

increasing the time-efficiency of GAs;

combined use of various heuristic opera-

tors (both existing and proposed) at the indi-

vidual level of interacting agents-processes for

the formation of new potential decisions (off-

spring-individuals);

adding internal iterations to the GA, pro-

viding the generation of a larger number of

offspring-individuals and potential decisions,

respectively;

providing the mechanism for periodi-

cally exchanging the best potential decisions

between agents-processes to avoid the jam-

ming problem of the GA at local extremes and

achieve an acceptable rate of population evolu-

tion for large-scale optimization problems.

An abstract description of the multi-agent

parallel genetic algorithm so developed is given

below.

Here,

i = 1, 2, ..., n – the index of decision variables

defining the values of the objective functions;

{ pi1, p

i 2 } – the pair of parent decision vari-

ables (parent-individuals) formed in the result

of the selection procedure (for example, using

tournament selection) for all i-ths decision

variables (i = 1, 2, ..., n);

– the pair of descendants (offspring-

individuals) formed by parents for all i-ths

decision variables (i = 1, 2, ..., n);

u (a, b), l (a, b), s (a, b) – random numbers

evenly distributed on the range of [a, b];

(k = 1, 2, ..., K ) – the index of parallel agent-

processes (GA), where K is the maximum

number of agent-processes in parallel GA;

gk

= 1, 2, ..., Gk – the index of internal itera-

tions belonging to the k-th agent-process.

The following new heuristic operators are

suggested for the real-coded genetic algorithm:

modified simulated binary crossover

(MSBX), provided the generation of potential

decisions in the continuous search space:

(1)



37

(2)

(3)

(4)

N = gk + 2, (5)

i = 1, 2, ..., n, k = 1, 2, ..., K, gk

= 1, 2, ..., Gk,

where , are coefficients (parameters of a

crossover), N is the parameter simulated the

number of bits in GAs with a binary coding

(N [2, Gk ]);

modified discrete SBX-crossover

(DMSBX), provided the generation of poten-

tial decisions in the discrete search space:

(6)

(7)

(8)

i = 1, 2, ..., n.

scalable uniform mutation operator

(SUM), provided quantizing of the feasible

ranges of decision variables into uniform inter-

vals to obtain potential solutions outside the

area of local extremes:

(9)

(10)

i = 1, 2, ..., n, gk

= 1, 2, ..., Gk, k = 1, 2, ..., K.

Note that all considered heuristic operators

are executed with a given probability. At the

same time, the probability of the execution of a

crossover operator at each iteration of the GA

is close to one (that is, the crossover is the most

important GA operator with real coding). The

probability of a mutation operator is selected

taking into account the relief of the objective

functions of the solved problem and, as a rule,

it is at the range [0.001, 0.1] while minimizing

the objective functions having relatively simple

relief, and in the range [0.1, 0.5] while minimiz-

ing complex objective functions with multiple

local extremes located near the global optimum.

Here,

tk

= 1, 2, ..., Tk – index of the external itera-

tions of the k-th agent process of GA, where Tk

is the number of external iterations;

gk = 1, 2, ..., G

k – the index of internal itera-

tions of the k-th agent process (GA), where Gk

is the number of internal iterations;

{LX, SBX, MSBX, DMSBX} – the set of pos-

sible crossover operators chosen with equal

probability at each tk-th step of GA, where LX

is the Laplas crossover, SBX – the standard

SBX-crossover, MSBX – the modified SBX-

crossover, DMSBX – the modified discrete

SBX-crossover;

{PM, UM, DUM, SUM} – the set of possible

mutation operators chosen with equal proba-

bility at each tk-th step of GA, where PM is the

power mutation operator, UM – the standard

operator of a uniform mutation, DUM – dis-

crete operator of a uniform mutation, SUM –

a scalable operator of a uniform mutation;

– the frequency of exchanging the best

potential decisions between all k-th process

agents (k = 1, 2, ..., K ).

Thus, the aggregated block diagram of the

proposed multi-agent parallel real-coded GA

developed for multi-objective optimization

(MA–RCGA–MO) can be presented in the

following form (Figure 1).

Note that the proposed GA is implemented

for the each parallel agent-process that peri-

odically exchanges the best (non-dominant

Pareto) potential decisions through the global

archive with all other agent- processes. Such

an approach can significantly increase the rate

of searching the Pareto-optimal solutions and

overcome the problem of a premature con-

vergence associated with frequent jamming

of GA at local extremes. Figure 2 shows the

aggregated architecture of the developed deci-

sion-making system in which the AnyLogic



38

Fig. 1. The block diagram of the developed multi-agent parallel real coded GA

Clearing the parent pool of potential decisions and the archive of non-dominated Pareto solutions.

Selection of a crossover operator from the set

Selection of a mutation operator from a set

Conducting tournament selection to form a pool of the most adapted parent-individuals.

Probabilistic selection of a pair of parents from the parent pool:

Execution of crossover and mutation operators to generate new potential decisions (offspring-individuals).

Computation of objective and fitness functions using AnyLogic for offspring-individuals:

Updating the population of the most adapted (non-dominant Pareto) individuals.

Updating the global population of the best (non-dominated) solutions, with a given frequency, i.e. if the following condition is performed:

Updating the local population of the -th agent-process by the best solutions from the global population with a given periodicity.

Stopping the GA when the required level of the rate of convergence is reached (the degree of stabilization of the fitness function values) for the global population.

YES

YES

NO

NO

Initialization of agent-process parameters (GA).

Generation of the initial population.

Calculation of the values of the objective functions in AnyLogic for each initial vector of decision variables.

The end



39

simulation models are integrated with the

developed genetic algorithm (MA–RCGA–

MO). The algorithm is implemented in C++

programming language using the MPI (Mes-

sage Passing Interface) technology, which

makes it possible to provide an efficient pro-

cedure of data exchange between all agents-

processes. Different software tools (for exam-

ple, GIS maps, graphs, tables, etc.) can be

used to provide the presentation of optimiza-

tion results previously written in the database

of the system (Oracle).

An important advantage of the suggested

architecture is the integration of the developed

parallel GA with AnyLogic simulation mod-

els (implemented on Java) using the JNI tech-

nology (Java Native Interface). Note that cur-

rently there are parallelization technologies for

the Java platform, for example, MPJ1, which

can also be applied to the AnyLogic models.

However, when solving large-scale optimiza-

tion problems, the most important factor is

the performance of the corresponding com-

putational procedures that can be significantly

improved only by using C++ and MPI tech-

nologies.

Fig. 2. The aggregated architecture of the decision-making systemfor socio-economic and ecological planning

1 http://www.mpjexpress.org/

JDBC

JDBC

Exporting the model as an executed JAR-file

AnyLogic simulation with implementation using Java

programming language (JAR file)

Oracle Call Interface (OCI LIB)

Runtime.getRuntime().exec()

JNI

Datasets for AnyLogic models, values of agent-processes parameters

and optimization results

Control Panel for the AnyLogic model

Multi-agent parallel genetic algorithmfor multi-objective optimization

(C++ and MPI)



40

2. An example of the practical implementation of the system for support

of decision-making

An example of a decision-making system

implemented for ecological and economic plan-

ning tasks related to the rational greening of the

city is considered further. Earlier we developed an

agent-based simulation model of the distribution

of harmful emissions in the city (in the AnyLogic

system) on the example of Yerevan, Republic

of Armenia [10]. Initially searching for the best

solutions in the model was conducted using a

parallel genetic algorithm with binary coding,

which required a lot of time for the generation

of a subset of the Pareto optimal solutions (sev-

eral hours of iterative calculations) on the server

HP ProLiant DL 380 GB with two 6 core proces-

sors Intel Xeon CPU E5645, 2.4 GHz and 64GB

of RAM, due to the large dimensionality of the

optimization problem being solved.

It should be noted that in the simulation model

two minimized objective functions were defined.

The first is the average daily pollution concen-

tration estimated in protected urban areas (in

particular, in the areas of kindergartens), as well

as the budget needed for greening the city to

ensure the natural protection of socially impor-

tant objects from harmful emissions produced by

enterprises and transport. At the same time, 111

kindergartens were previously selected for pro-

tection by trees at the individual level, taking into

account the variability of such parameters as the

type of trees (for example, poplar, maple, oak,

spruce, elm), the distance between the clusters

of trees (from 5 to 60 meters), the radius of the

planting zone (from 30 to 100 meters) and the

geometry of planting trees around kindergartens

(for example, a simple circle, an arithmetic spi-

ral, a double circle, etc.).

In the result of the application of the devel-

oped multi-agent parallel real-coded genetic

algorithm (MA–RCGA–MO), the time-effi-

ciency of the search procedure for optimum

solutions to the considered problem of complex

ecological and economic system of the city was

significantly improved. Using MA–RCGA–

MO, the best scenario was found for a polyno-

mial time, providing almost fourfold reduction

in the concentration of harmful emissions in

the atmosphere in protected urban areas with

an acceptable level of greening expenses. The

optimization results, previously saved on the

Oracle DBMS, were visualized on the map of

Yerevan using the AnyLogic system (Figure 3).

Conclusion

This paper presents a new multi-agent paral-

lel real-coded genetic algorithm MA–RCGA–

MO, which provides an effective procedure for

finding Pareto optimal solutions in large-scale

multi-objective optimization problems.

An important feature of the suggested genetic

algorithm is use of new heuristic crossover and

mutation operators, the characteristics of which

functionally depend on the number of the associ-

ated agent-process, as well as providing a mech-

anism for periodic exchange of the best potential

decisions between all agents-processes to avoid

a premature convergence (caused by potential

jamming the GA at local extremes) and increase

the rate of search for optimal solutions.

The important advantage of the suggested

multi-agent GA is its aggregation with the

AnyLogic simulation models through objective

functions. At the same time, C++ program-

ming language and MPI technology provide an

effective procedure for periodically exchanging

the best potential decisions, and the JNI tech-

nology provides the ability to integrate the GA

with the AnyLogic models.

In future work, it is planned to implement dif-

ferent approaches to the generation of Pareto

optimal solutions (for example, NSGA-III) for

the developed multi-agent parallel genetic algo-

rithm with studies of the effectiveness of appro-

priate modifications. Moreover, we expect the

implementation of multi-agent genetic algo-

rithm using technology CUDA (Compute Uni-

fied Device Architecture).



41

Fig.

3. V

isua

lizat

ion

of th

e re

sults

of m

inim

izatio

n of

har

mfu

l em

issi

ons

in th

e ci

ty u

sing

the

prop

osed

gen

etic

alg

orith

m



42

References

1. Khivintcev M.A., Akopov A.S. (2014) Application of multi-agent genetic algorithm for search

of optimum strategic and operational decisions. Business Informatics, no 1 (27), pp. 23–33

(in Russian).

2. Khivintcev M.A., Akopov A.S. (2013) Distributed evolutionary network for the solution of

multi-objective optimizing problems in simulation systems. Business Informatics, no 3 (25),

pp. 34–40 (in Russian).

3. Akopov A.S., Hevencev M.A. (2013) A multi-agent genetic algorithm for multi-objective

optimization. Proceedings of the 2013 IEEE International Conference on Systems, Man and

Cybernetics. Manchester, UK, 13–16 October 2013, pp. 1391–1395.

4. Bleuer S., Brack M., Thiele L., Zitzler E. (2001) Multiobjective genetic programming:

Reducing bloat by using SPEA 2. Proceedings of the 2001 Congress on Evolutionary

Computation (CES–2001). Seoul, South Korea, 27–30 May 2001, vol. 1, pp. 536–543.

5. Akopov A.S., Beklaryan L.A., Saghatelyan A.K. (2017) Agent-based modelling for ecological

economics: A case study of the Republic of Armenia. Ecological Modelling, vol. 346,

pp. 99–118.

6. Akopov A.S., Beklaryan A.L., Khachatryan N.K., Fomin A.V. (2017) The forecasting system

dynamics of oil production using simulation modeling. Information Technologies, vol. 23, no 6,


7. Akopov A.S., Beklaryan A.L., Khachatryan N.K., Fomin A.V. (2018) Development of an adap-

tive genetic optimization algorithm using agent modeling methods]. Information

Technologies, vol. 24, no 5, pp. 321–329 (in Russian).

8. Herrera F., Lozano M., Verdegay J.L. (1998) Tackling real-coded genetic algorithms: operators

and tools for the behaviour analysis. Artificial Intelligence Review, vol. 12, no 4, pp. 265–319.

9. Akopov A.S. (2014) Simulation modeling. Moscow: Urait (in Russian).

10. Akopov A.S., Beklaryan G.L., Beklaryan L.A. (2017) Agent-based modelling for ecological-

economics system of a city (on example of Yerevan, Republic of Armenia). Artificial Societies,

vol. 12, no 3-4. Available at: https://artsoc.jes.su/s207751800000109-8-1 (accessed 01

September 2018) (in Russian).

11. Akopov A.S., Beklaryan A.L., Beklaryan L.A., Saghatelyan A.K. (2016) Modelirovanie

regional’noj ehkologo-ehkonomicheskoj sistemy s mekhanizmom gosudarstvennogo

regulirovaniya na primere Respubliki Аrmeniya [Modelling the regional ecological-economic

system with the mechanism of the government regulation for the case study

of the Republic of Armenia]. Economics of Contemporary Russia, vol. 72, no 1,


Acknowledgements

This work was supported by the Russian Foun-

dation for Basic Research (grant No. 18-51-

45001), as well as the Department of Science

and Technology, Government of India (project

No. INT/RUS/RFBR/305).



43

12. Beklaryan L.A., Akopov A.S., Beklaryan A.L., Saghatelyan A.K. (2016) Agent-based

simulation modelling for regional ecological-economic systems. A case study of the Republic

of Armenia. Journal of Machine Learning and Data Analysis, vol. 2, no 1, pp. 104–115.

13. Akopov A.S., Beklaryan L.A., Beklaryan A.L., Saghatelyan A.K. (2014) The integrated model

of eco-economic system on the example of the Republic of Armenia. Computer Research and

Modeling, vol. 6, no 4, pp. 621–631 (in Russian).

14. Akopov A.S. (2014) Parallel genetic algorithm with fading selection. International Journal

of Computer Applications in Technology, vol. 49, no 3/4, pp. 325–331.

15. Akopov A.S. (2012) Designing of integrated system-dynamics models for an oil company.

International Journal of Computer Applications in Technology, vol. 45, no 4, pp. 220–230.

16. Beklaryan A.L., Akopov A.S. (2018) Simulation model of the optimal allocation of credit

applications for interregional underwriting center of a commercial bank. Vestnik komp’iuternykh

i informatsionnykh tekhnologii, no 11, pp. 45–56 (in Russian).

17. Akopov A.S., Beklaryan L.A. (2015) An agent model of crowd behavior in emergencies.

Automation and Remote Control, vol. 76, no 10, pp. 1817–1827.

18. Beklaryan A.L., Akopov A.S. (2016) Simulation of agent-rescuer behaviour in emergency

based on modified fuzzy clustering. Proceedings of the 2016 International Conference

on Autonomous Agents and Multiagent Systems (AAMAS 2016). Singapore, 9–13 May 2016,

pp. 1275–1276.

19. Zhang Q., Liu W., Li H. (2009) The performance of a new version of MOEA/D on CEC09

unconstrained MOP test instances. Proceedings of the 2009 IEEE Congress on Evolutionary

Computation (CEC 2009). Trondheim, Norway, 18–21 May 2009, pp. 203–208.

20. Deb K., Pratap A., Agarwal S., Meyarivan T. (2000) A fast elitist multi-objective genetic

algorithm: NSGA-II. IEEE Transactions on Evolutionary Computation, vol. 6, no 2,

pp. 182–197.

21. Jain H., Deb K. (2014) An evolutionary many-objective optimization algorithm using

reference-point based nondominated sorting approach. Part II: Handling constraints

and extending to an adaptive approach. IEEE Transactions on Evolutionary Computation,

vol. 18, no 4, pp. 602–622.

22. Zhong W, Liu J, Xue M., Jiao L. (2004) A multiagent genetic algorithm for global numerical

optimization. IEEE Transactions on Systems, Man, and Cybernetics. Part B (Cybernetics),

vol. 34, no 2, pp. 1128–1141.

23. Deb K., Kumar A. (1995) Realcoded genetic algorithms with simulated binary crossover:

Studies on multimodal and multiobjective problems. Complex Systems, vol. 9, no 6,

pp. 431–454.

24. Deep K., Thakur M. (2007) A new crossover operator for real coded genetic algorithms.

Applied Mathematics and Computation, vol. 188, no 1, pp. 895–911.

25. Deep K., Thakur M. (2007) A new mutation operator for real coded genetic algorithms.

Applied Mathematics and Computation, vol. 193, no 1, pp. 211–230.



44

About the authors

Andranik S. Akopov

Dr. Sci. (Tech.);

Professor, Department of Business Analytics, National Research University Higher School of Economics,

20, Myasnitskaya Street, Moscow 101000, Russia;

Chief Researcher, Laboratory of Dynamic Models of Economy and Optimization,

Central Economics and Mathematics Institute, Russian Academy of Sciences,

47, Nakhimovky Prospect, Moscow 117418, Russia;


Armen L. Beklaryan

Cand. Sci. (Tech.);

Associate Professor, Department of Business Analytics, National Research University Higher School

of Economics, 20, Myasnitskaya Street, Moscow 101000, Russia;

Senior Researcher, Laboratory of Social Modeling, Central Economics and Mathematics

Institute, Russian Academy of Sciences, 47, Nakhimovky Prospect, Moscow 117418, Russia;


Manoj Thakur

PhD;

Associate Professor, School of Basic Sciences, Indian Institute of Technology Mandi, Mandi, Himachal

Pradesh 175005, India;


Bhisham Dev Verma

Doctoral Student, School of Basic Sciences, Indian Institute of Technology Mandi, Mandi, Himachal

Pradesh 175005, India;




45

Development of strategic management tools for heat supply enterprises in the Donetsk region

Mariya A. Myznikova a


Larisa N. Brazhnikov b


a Donetsk National University Address: 24, Universitetskaya Street, Donetsk 283001, Ukraineb Academy of Economic Sciences of Ukraine Address: 2, Zhelyabova Street, Kiev 03057, Ukraine

Abstract

Raising the eff ectiveness of strategic management in conditions of high complexity and dynamic change of modern management systems requires the development of an appropriate mathematical toolkit. The task of raising eff ectiveness of strategic management is especially topical for heat supply enterprises of the Donetsk region, where operations have been complicated by a number of general system problems, and by the presence of substantial external challenges. At the same time, the question of using mathematical apparatus to raise the eff ectiveness of strategic management of enterprises in the sphere of residential-communal services appears not to have been widely studied. In this regard, the objective of this study is raising the eff ectiveness of strategic management of heat supply enterprises of the Donetsk region by developing a respective toolkit of mathematical modeling. To achieve the goal we have set, in this work we carried out an analysis of the viability of the system using the methodology proposed by S. Beer; we made an analysis of the elements of the market of heat supply, and also developed system dynamic models based on the approach of J.W. Forrester.

As a result of our research, we discovered the basic problems infl uencing the viability of the system at the strategic level. It was established that the problems revealed are the consequence of the imperfections of the methodological base, including absence of timely information on the dynamics of the external environment, forecasting of the key parameters, a toolkit for making decisions, etc. For the purpose of fi nding a toolkit to improve the methodological base, we performed an analysis and forecast of the heat supply market in the Donetsk region as part of the external environment which exerts a very signifi cant infl uence on the activity of the heat supply enterprises of the Donetsk region.

In the course of this market analysis, we established that the off er of heat supply services is not constant and depends on the tariff setting costs. Due to this, we proposed an approach to forecasting tariff setting costs based on the methodology of A.G. Ivakhnenko but distinguished from that by the



46

Introduction

The complexity of contemporary eco-

nomic systems resulting from the

high agility of the processes occur-

ring in them, by the presence of external chal-

lenges, by the large quantity and non-linearity

of links of elements in such systems, leads to

reduced effectiveness of traditional methods of

management and makes it very topical to use

a complex approach based on the toolkit of

mathematical modeling.

The desirability of raising the effectiveness

of strategic management by developing a com-

plex toolkit acquires special relevance for heat

supply enterprises of the Donetsk region, due

to several factors: the primary importance of

heat supply services to protect the lives of the

population and ensure the functioning of the

region’s companies, as well as the scale of the

regional market of residential communal ser-

vices (including heat supply services). Thus,

the residential communal entities of the region

cover the needs of around 2.5 million people, as

well as more than 180 major enterprises. More

than 8% of the working population is engaged

in this sphere.

Furthermore, the problems with which

heat supply enterprises of the Donetsk region

encounter bear a deep and general systemic

character. These include a high degree of dete-

rioration of the basic assets, insufficient vol-

umes of financing, low solvency of the con-

sumers, large amounts of receivables, etc.

Solutions to the enumerated problems

require high effectiveness of strategic manage-

ment, which is difficult to achieve when the

respective toolkit is missing.

A whole series of both foreign and domestic

works has been devoted to the issues of find-

ing more effective approaches to management

(including strategic management) of heat sup-

ply enterprises, among them by E.Yu. Adzh-

agulov [1], D.L. Bakieva [2], E.V. Baland-

ina [3], Е.Е. Vorobieva [4], А.V. Darbasov

[5], Т.А. Makarenya [6]. Such scholars as

А.V. Allakhverdyan [7], L.N. Brazhnikova

[8, 9], S.G. Kulikov [10], R.N. Lepa [11],

Ya.A. Lyashok [12], E.A. Perkova [13],

V.P. Poluyanov [14], and I.A. Yurchenko [15]

are among those who have dedicated their

works to the specifics of heat supply companies

of the Donetsk region.

At the same time, despite the attention

researchers have given to questions of the

effectiveness of management of companies in

the residential communal sphere the problem

presence of a training sample and two test samples. In addition, in the course of analyzing the market we discovered new forms of demand for heat supply services – lost demand and unpaid demand. On the basis of the dependencies established, we built a model for forecasting the behavior of consumers of a heat supply company oriented to the level of marketing. With the help of this model, by means of supplements to it and modifi cations, we built a complex model of strategic management of heat supply enterprises of the Donetsk region allowing us to analyze the eff ectiveness of using one or another lever of strategic management on the basis of scenario analysis.

Key words: strategic management; system dynamic modeling; sustainability; tariff setting costs;

heat supply enterprise.

Citation: Myznikova M.A., Brazhnikova L.N. (2019) Development of strategic management tools

for heat supply enterprises in the Donetsk region. Business Informatics, vol. 13, no 1, pp. 45–58.

DOI: 10.17323/1998-0663.2019.1.45.58



47

of developing an effective toolkit of strategic

management of heat supply enterprises remains

unresolved. Thus, works devoted to application

of the apparatus of economic-mathematical

modeling and modern tools for raising effec-

tiveness of strategic management in the sphere

of residential communal services have focused

on solving the problems at the regional and

municipal levels. Meanwhile, the questions of

developing a toolkit for mathematical mode-

ling to raise the effectiveness of strategic deci-

sions at the level of management of individual

heat supply enterprises remain among the least

developed, all of which predetermine the time-

liness of the line of research we have chosen.

In connection with the foregoing, the objec-

tive of this research is to raise the effectiveness

of strategic management of heat supply enter-

prises of the Donetsk region by developing a

respective toolkit of mathematical modeling.

In accordance with the goal of the research,

we formulated a methodological base which is

comprised of the works of a number of authors

devoted to the questions of raising effectiveness

of management of business systems by using

mathematical apparatus [16–27].

1. Analysis of the viability

of a heat supply enterprise

of the Donetsk region

Raising the effectiveness of strategic manage-

ment of a heat supply enterprise of the Donetsk

region requires that we carry out a retrospec-

tive analysis of the functioning of the system, as

well as that we reveal and systematize the basic

problems of enterprises in the given sphere.

For this, in the framework of a heat supply sys-

tem, we can distinguish six subsystems relating

to various levels of management - strategic, tac-

tical and operative. In particular, at the strategic

level we find the subsystem for decision making

and the subsystem providing information. At the

tactical level, we find the subsystem for distribu-

tion of resources in short supply, the subsystem

of internal audit and the subsystem for resolving

current problems. At the operative level, there is

the “operational element” subsystem.

Let us examine the characteristics of the enu-

merated subsystems and the causes of problems

arising inherent in the current state of the heat

supply system.

The subsystem for decision making is char-

acterized by the fact that the basic functions

of decision making relating to the entire heat

supply system are accorded to the bodies of

departmental control. The reasons why prob-

lems arise are the economic and social ground-

lessness of the tariffs, as well as the imperfec-

tions of the methodological base which enables

one to react to deviations which arise.

The subsystem of information manage-

ment is supposed to send to the decision-mak-

ing subsystem information on the state of the

external world and basic trends of its change, as

well as information about the necessary action

in response to these changes. At the same

time, the de facto fulfillment by the informa-

tion management subsystem of its functions

is limited exclusively to stating the actual val-

ues over the preceding periods. The sources of

the problems – the lack of up-to-date infor-

mation about the functioning of the real sec-

tor of the economy, significant time lags, and

also the lack of formalization of threshold val-

ues of deviations below which it is necessary to

take control.

The function of the subsystem for distrib-

uting resources in short supply in the system

of heat supply is carried out by various insti-

tutions: distribution of subsidies is provided

from local budgets, other subsidies come from

the budget of the republic, investments from

all subjects of the market. Moreover, the dis-

tribution of investment flows at the micro

level is performed by the operational elements

independently. The sources of the problems –

ineffectiveness of the distribution of resources

in short supply, and also growth of expenses

for the bureaucratization of processes of issu-



48

ing subsidies, financing and regulating the

system.

To provide for the functions of internal audit,

there are specialized commercial organiza-

tions whose services are paid by the enterprise.

However, the selection of these organizations

is done, as a rule, by local self-government

authorities. The causes of the problems exist-

ing in this area arise beyond the boundaries of

the given research.

The subsystem for resolving current problems

is represented by laws and sublegal documents

regulating the activity of enterprises from the

sectoral ministry and local authorities of self-

administration. Here among the sources of the

problems we can mention the legally nonreg-

ulated nature of the monopoly status of the

enterprises, the forms and methods of republic

tariff policy, as well as the forms and methods

for recovering accounts receivable.

Finally, the “operative element” subsystem is

represented in the form of enterprises of various

kinds of ownership and functional affiliation.

Here the causes of the problems which arise are

the deteriorated state of the basic assets, the use

of outdated technologies, high expenses and

low efficiency, an ineffective innovation and

investment policy, the unsatisfactory financial

condition of the enterprise, an ineffective pol-

icy on price formation, an ineffective system

of managing expenses, an imperfect system

of managing the receivables and credit policy,

the low quality of services provided, and low

attractiveness for investment.

Due to the specific external conditions of the

functioning of heat supply enterprises of the

Donetsk region (as with many other systems in

depressed territories) one of the most impor-

tant tasks of their operations is to achieve a sys-

tem of viability, which is taken to mean “the

ability of the system to independently support

its autonomous existence for as long as pos-

sible” [16]. This characteristic of the system

has been given the name “viability” and was

described in the works of S. Beer [16], as well

as by a broad circle of scholars in the context of

systemic and cyber approaches.

In connection with the foregoing, it is inter-

esting to analyze the problems of the function-

ing of heat supply enterprises of the Donetsk

region at various levels of a viable system. The

basic problems of strategic management of

a viable system of heat supply are related to

shortcomings of the subsystems of information

management and management decision mak-

ing.

These problems are the consequence of

shortcomings of the methodological base,

including lack of up-to-date information about

the dynamics of the external environment, an

ineffective approach to forecasting key param-

eters, as well as lack of a toolkit to support the

adoption of strategic decisions.

Due to the fact that one of the key prob-

lems of managing heat supply enterprises

of the Donetsk region is lack of up-to-date

information on the dynamics of the exter-

nal environment, and also proceeding from

the goals of raising the effectiveness of strate-

gic management of heat supply enterprises of

the Donetsk region, it is worthwhile to do an

analysis of and forecast of the external envi-

ronment. Moreover, in the context of a cyber

approach it is customary to separate out the

parts of the external environment which exert

the most significant impact on the subject of

the research (external supplement [16]). For

the enterprises we analyzed, the market of

heat supply in the Donetsk region can be seen

as the external supplement.

2. Analysis of the market elements

of heat supply: demand, supply, price level,

market conditions

We take the heat supply market to mean

the exchanges which develop between its par-

ticipants based on the sale-purchase of spe-

cific benefits (hot water supply and provision

of heating) which can be measured quantita-



49

tively and described in the form of characteris-

tic behavior of market participants.

Analysis of the specialized literature made it

possible to reveal the basic common elements

of the market for residential communal ser-

vices. They comprise: offer of services by the

residential-communal companies, demand for

residential-communal services, tariffs for res-

idential-communal services and the market

conditions.

Given the limitation and complexity of influ-

ence on the market conditions, what is of prac-

tical interest for the purposes of management is

to review the categories of demand and supply

on the market of residential-communal ser-

vices, as well as the processes of price forma-

tion.

Development of the model basis for evalu-

ating the behavior of producers of heat sup-

ply services (i.e. evaluation of the offer) has

been described in the work [28]. In particu-

lar, it was established that offer by heat sup-

ply enterprises in the Donetsk region is not a

fixed amount, i.e., it changes under the influ-

ence of several factors. At the same time it

was shown that the use of price (tariffs) as the

main factor of the offer with respect to heat

supply services is wrong, insofar as the tariff

is a conditional value. In this connection, we

proposed to use as the basic factor tariff set-

ting costs. The dependency shown raises the

relevance of applying up-to-date methods of

forecasting tariff setting expenses of heat sup-

ply enterprises of the Donetsk region. For

determination of the character and closeness

of the bond, we analyzed the existing meth-

ods of selecting parameters and established

that their application is difficult under condi-

tions of the economic shocks which the econ-

omy of the Donetsk region has been experi-

encing ever since 2014. In this respect, in [28]

we see further development of the inductive

method of organizing models of complex sys-

tems proposed by A.G. Ivakhnenko [18]. The

approach to constructing a model for forecast-

ing expenses underlying the tariffs of heat sup-

ply enterprises is distinguished by the exist-

ence of training and two test samples, all of

which allows us to analyze the suitability of the

basic model under conditions of the economic

shocks of 2014–2015, and also the suitability of

the refined model for forecasting the following

trend of the readings.

Analysis of the system of tariff formation and

its interdependence with consumer behav-

ior (demand) in heat supply enterprises of the

Donetsk region allowed us to establish that low

solvency of the population causes a high level

of accounts receivable and serves as an impedi-

ment to establishment of economically justified

tariffs. In this connection, in [29] a method is

proposed for calculating the critical maximum

tariffs for services of residential-communal

service enterprises.

Study of the behavior of consumers (demand)

is described in the works [30, 31], where it is

proposed to examine lost demand, which is

expressed as refusal to buy services of central-

ized heat supply and as unpaid demand, which

takes the form of consumer debt (accounts

receivable).

Revealing the factors influencing the dynam-

ics of market elements of heat supply and

behavior of its subjects lay at the basis of our

building system dynamic modeling of strategic

management.

3. Constructing system dynamic models

of strategic management of a heat supply

enterprise of the Donetsk region

An original toolkit for system dynamic mod-

eling of strategic management of a heat sup-

ply enterprise consists of two basic models: an

simulation model of forecasting the behavior of

the enterprise’s consumers [32] and a complex

model of strategic management of a heat sup-

ply enterprise.

It should be noted that the simulation model

of forecasting the behavior of a heat supply



50

enterprise’s consumers is oriented at the level

of marketing and therefore has a number of

limitations and simplifications. The complex


ply enterprise is based on a model of forecast-

ing the behavior of consumers but is geared to

the level of top leadership of the enterprise, and

therefore presupposes a larger number of man-

agement levers, as well as a smaller number of

limitations and simplifications.

3.1. The simulation model of forecasting the behavior of a heat supply

enterprise’s consumers

Changes in the behavior of consumers exert

an influence on the financial results of the

activities of heat supply enterprises. More-

over, it should be noted that rejection of the

services of heat supply enterprises exerts an

influence on the functioning of the enterprise

both in the short term and as regards long

term prospects.

Under conditions in which the heat sup-

ply enterprises of the Donetsk region operate,

consumer arrears (accounts receivable) have a

short term influence on the financial perfor-

mance of the enterprise, and also have a per-

sistent tendency to turn into hopeless, frozen

assets for the long term, changing the nature

of this influence from short term to long term.

Thus, one may conclude that an increase in

lost demand, as well as the growth in unpaid

demand exert an influence both on short term

and on long term financial results from opera-

tions in heat supply enterprise. It follows from

this that research into the behavior of consum-

ers has high theoretical and practical signifi-

cance for developing a toolkit to raise the effec-

tiveness of strategic management.

On the basis of the analysis carried out, the

equation of dependency of indicators of the

increased level of accounts receivable from the

population on the correlation of the level of

tariffs and the level of salaries can be presented

in the following manner:

(1)

where a01

, a11

– regression coefficients of the

model;

– the level of tariffs for the population of

heat supply services on the heat supply market

at moment in time t;

M t – the average level of salaries in the region

at moment in time t.

At the same time, the interest of enterprises,

unlike the population, is formed under the

influence of two factors – the quality of the ser-

vices offered and the system of material incen-

tives. There is practically no system of mate-

rial incentives in the heat supply market of the

Donetsk region. Due to this, it is worthwhile

reviewing the dependence of increased level of

consumer arrears (accounts receivable) on the

quality of services provided.

Tariffs exert an influence on the possibil-

ity and ability to pay for heat supply services.

Thus, growth in accounts receivable of heat

supply companies may be presented in the fol-

lowing way:

(2)

where – growth in the level of accounts

receivable of the k category of consumers for

the period [t0; t];

k – categories of consumers, k [1; 4];

W t – the quality of heat supply services at

moment in time t;

a0k

, a1k

, a2k

– regression coefficients of the

model.

Lost demand is also an indicator which to a

certain extent depends on the level of tariffs

and quality of services. Quality of heat sup-

ply services is an aggregate indicator which is

calculated from information about the quality

of boilers, networks and communications, as

well as the quality of the accompanying ser-

vice.



51

Evaluation of the quality of the service for

heat supply services is made using a calculation

of an integral reading which changes within

a given range [0; 10] and is calculated using

the sum of partial coefficients determined by

expert evaluations.

The model obtained for forecasting the

behavior of consumers of a heat supply enter-

prise in its most simplified form can be pre-

sented in the form of a diagram of cause and

effect links (Figure 1).

3.2. The complex model of strategic

management of a heat supply enterprise

On the basis of our analysis of the model of

forecasting the behavior of heat supply enter-

prise consumers, one can conclude that the

problems of managing the behavior of the

enterprise’s consumers are in direct depend-

ency relations with effectiveness of the pol-

icy of tariff formation, as well as policy in the

sphere of quality of services provided.

At the same time, based on the orientation

of the model towards the level of the market-

ing department, it presupposes the following

assumptions and simplifications:

a change in the level of tariffs, as well as in

the level of quality over time, and not under the

influence of management decisions;

it does not presuppose analysis of the cost

and effectiveness of such management deci-

sions;

Accounts receivable

(population)

Accounts receivable

(enterprises)

Accounts receivable

(state owned entities)

Quality of boilers

Heat sales

Quality of final services

Average salary

Tariff setting costs

Current tariffs

Income received in liquid form

Lost income

Lost demand

Tariff level for

enterprises

Tariff level for

population

Level of dismantled personal accounts

Quality of networks and

communications

Quality of accompanying

service

Fig. 1. Model of forecasting the behavior of heat supply enterprises’ consumers in the form of causal links

Accounts receivable

Accounts receivable (municipal

entities)

Income



52

it does not consider the investment policy;

the model does not allow us to trace the

influence of the indicators being analyzed on

the financial results of the enterprise’s activi-

ties.

In this connection, and also to improve the

methodology of managing a heat supply enter-

prise as a whole, it is interesting to see mod-

ification of the previously developed model

of forecasting the behavior of the enterprise’s

consumers by changing it and supplementing it

to reach a complex model of strategic manage-

ment of a heat supply enterprise.

Based on the assumptions and simplifica-

tions described above, it is proposed to make

the following changes in the complex model of

strategic management:

to establish the mutual dependency between

the level of tariffs and the quality of services, as

well as with the management decisions adopted;

analysis of the elasticity of the level of

quality and service in relation to the expenses

needed to raise it;

analysis of the effectiveness of manage-

ment decisions and, in particular, decisions on

the investment policy;

inclusion in the model of the indicator of

financial results from the enterprise’s activities

as a resulting element.

In this regard, in the complex model the qual-

ity of boilers, equipment, networks and com-

munications depends on the total investments

of the enterprise, with a delay in two periods

(quarters), and the level of service depends on

the total investments or the enterprise in the

current period.

The effectiveness of the management deci-

sions adopted is determined according to the

formula:

(3)

where – the effectiveness of the m-th man-

agement decision at moment in time t;

– the financial result (profit/loss) at

moment in time t under conditions of the

implementation of the m-th management

decision;

– the financial result (profit/loss) at

moment in time t under conditions of the

implementation of the basic scenario (keeping

the current dynamics of the indicators);

– total expenses of the heat supply enter-

prise at moment in time t under conditions of

the implementation of the m-th management

decision;

– total expenses of the heat supply enter-

prise at the moment in time t under conditions

of implementation of the basic scenario (keep-

ing the current dynamics of the indicators).

It should be mentioned that the calcula-

tion of the financial result at moment in time t

presupposes inclusion of the accounts receiv-

able. At the same time, the accounts receiv-

able of heat supply enterprises of the Donetsk

region have very low liquidity. Due to this, we

introduced the term “absolutely liquid finan-

cial result,” meaning the financial result of

a heat supply enterprise without considering

accounts receivable. Thus, there is enhanced

interest in the calculation of the indica-

tor of effectiveness of management decisions

expressed in the increment of absolutely liq-

uid financial results:

(4)

where – the effectiveness of the m-th man-

agement decision expressed as the increase of

absolutely liquid financial results at moment in

time t (liquid effectiveness);

– absolutely liquid financial results

(profit/loss) at moment in time t under condi-

tions of the implementation of the m-th man-

agement decision;

– absolutely liquid financial results

(profit/loss) at moment in time t under condi-

tions of the implementation of the basic sce-



53

nario (maintaining the current dynamic of the

indicators).

Thus, liquid effectiveness is the liquid effect

correlated with costs for achieving it.

On the basis of the foregoing, and also on

the basis of the previously described model

of forecasting the behavior of consumers of a

heat supply enterprise we developed a complex


ply enterprise of the Donetsk region. The main

elements and interconnections of the model in

the form of a diagram of cause and effect are set

out in Figure 2.

By share of dismantled personal accounts we

mean the correlation of the number of disman-

tled personal accounts (i.e., by those who have

rejected using the services of centralized heat-

ing) and the overall share of personal accounts

served by the heat supply enterprise.

By element of the model of ‘aggregated qual-

ity’ we mean the indicator reflecting the overall

level of quality of heat supply services, includ-

ing the quality of the boilers and networks, as

well as the quality of the accompanying ser-

vice. The numerical values of the indicator are

obtained by polling the users of the services.

The model we developed allows us to analyze

the effectiveness of one or another lever for pur-

poses of obtaining economic results expressed

as the change of the financial result or of the

absolutely liquid financial result of the activi-

ties of heat supply enterprises of the Donetsk

region.

The following are used as management levers

(managing the parameters) in the model:

the level of tariffs for enterprises;

the level of tariffs for the population;

the volume of investments in moderniza-

tion of the networks;

the volume of investments in moderniza-

tion of the boilers;

the volume of investments in improving the

service;

the volume of other costs.

Thus, the model allows us to raise the effec-

tiveness of the investment policy, the tariff pol-

icy and the policy in the sphere of managing

costs.

The block of evaluation of the effectiveness

incorporated in the model is intended for cal-

culating the effectiveness of management deci-

sions expressed as the incremental growth of

both the financial result and the absolutely liq-

uid financial result.

As a constant we used the discount rate and

the volume of state investments. All other indi-

cators of the model are calculated and obtained

by applying a modification of the inductive

method of self-organization of the models of

complex systems, the method of lowest quad-

rates, spline-interpolation, etc.

3.3. Results

of simulation modeling

On the basis of the complex model of stra-

tegic management of heat supply enterprises

described above, using scenario analysis, we

carried out a series of experiments allowing us

to determine the most effective management

levers. The results of the numerical experi-

ments are shown in Figure 3.

For evaluation of the effectiveness of apply-

ing various management levers we carried out

the following experiments:

Scenario 1: Increasing the level of tariffs for

enterprises by 10%;

Scenario 2: Increasing the level of tariffs for

the population by 10%;

Scenario 3: Increasing the volume of invest-

ments for modernization of the networks by

10%;


ments for modernization of the boilers by 10%;


ments for improving the service by 10%.

The results of the numerical experiments are



54

Fig. 2. The complex model of strategic management of a heat supply enterprise in the form of a causality diagram


Indi

cato

r of t

he

spee

d of

spe

cific

in

crem

ent i

n va

lue

Inco

me

rece

ived

in

liqui

d fo

rm

Net p

rese

nt

valu

e

Liqu

id

effic

ienc

y

Effic

ienc

y

Effe

ct

Liqu

id

finan

cial

resu

lt

Pote

ntia

l inc

ome

Inco

me

Fina

ncia

l res

ult

Tota

l oth

er c

ostsLo

st in

com

e

Net f

inan

cial

resu

lt

Acco

unts

rece

ivab

le (e

ntiti

es

finan

ced

from

loca

l bud

gets

)Ac

coun

ts re

ceiv

able

(b

udge

t ent

erpr

ises

) Ac

coun

ts re

ceiv

able

(o

ther

con

sum

ers)

Acco

unts

rece

ivab

le

(pop

ulat

ion)

Lost

dem

and

Shar

e of

dis

man

tled

pers

onal

acc

ount

s

Qual

ity o

f ac

com

pany

ing

serv

ice

Unpa

id d

eman

d

Aver

age

sala

ry

Shar

e of

em

erge

ncy

com

mun

icat

ions

Shar

e of

em

erge

ncy

boile

rs

Othe

r cos

ts

Inve

stm

ents

Tarif

f to

aver

age

sala

ry ra

tio

Tarif

f for

en

terp

rises

Tarif

f for

po

pula

tion

Tota

l tar

iff s

ettin

g co

sts

Aggr

egat

ed q

ualit

y

Wei

ghte

d av

erag

e ta

riff

Taxe

s an

d fe

es

Tota

l cos

ts

Qual

ity o

f boi

lers

and

co

mm

unic

atio

ns

Liqu

id e

ffect

Disc

ount

rate


55

compared with the forecast values of the opera-

tions of the system (Scenario 0).

Due to the fact that reducing other costs in

the long-term perspective is a precondition for

lowering the quality of services, the given man-

agement lever cannot be viewed as an effective

instrument of strategic management.

The economic effect expressed in the change

of financial result obtained taking into account

implementation of various scenarios is shown

in Table 1. We note that the first step of mod-

eling corresponds to the first quarter. Insofar

as the model reflects both retrospective data

which cannot be changed by implementation

of the scenarios (modeling steps 1–19) and

the forecast values of indicators assuming the

implementation of one of the scenarios (steps

20–30), the evaluation of the economic effect

is seen as worthwhile beginning only from the

20-th period of modeling.

The graphic illustration of calculated val-

ues of the economic effect obtained expressed

in the change of financial result following the

implementation of the indicated scenarios is

shown in Figure 4.

As we see in Figure 4, implementation of sce-

nario 5 is the most justified, i.e., increasing the

volume of investments in improving the service.

Conclusion

Thus, we propose an approach to increasing

the effectiveness of strategic management of

heat supply enterprises of the Donetsk region

based on the development of a respective

toolkit of mathematical modeling. Use of the

proposed approach assumes step-by-step solu-

tion of a number of tasks, namely:

analysis of the viability of the object of stra-

tegic management with a view to revealing the

most significant problems which bear on the

ability of the system independently to main-

tain its autonomous existence;

analysis of the market elements as the most

significant part of the external environment

where the enterprise operates;

development of approaches to raising the

effectiveness of forecasting the behavior of sub-

jects of the market as the methodological basis

of the system of information management serv-

ing the strategic management of the enterprise;

Fina

ncia

l res

ult,

‘000

rub.

Fina

ncia

l res

ult,

‘000

rub.

Fina

ncia

l res

ult,

‘000

rub.

Fina

ncia

l res

ult,

‘000

rub.

Fina

ncia

l res

ult,

‘000

rub.

Fina

ncia

l res

ult,

‘000

rub.

Fig. 3. The results of numerical experiments with the complex model of strategic management of heat supply enterprises using PowerSim software

Time, quarters

Time, quarters

Time, quarters

Time, quarters

Time, quarters

Time, quarters



56

building a model basis to support decision

making, taking into account the main correla-

tions we discovered and allowing us to perform a

scenario analysis of the effectiveness of manage-

ment levers;

carrying out numerical experiments with

the model, as a result of which we established

that the most effective management lever is

raising the volume of investments directed into

improvements to the accompanying service.

Use of the given lever allows us to receive an

economic effect of 17,071,830.7 rubles in the

first quarter. Moreover, we forecast a growth of

the economic effect given systematic applica-

tion of this management lever.

As regards the direction of further research,

we can mention adaptation of the results

obtained to a wide range of objects, as well as

the programmatic realization of a system for

supporting the decisions taken.

Table 1.Calculation of the economic effect expressed in change

of the financial result obtained as a result of implementing the scenarios, ‘000 rub.

Step of modeling, quarter Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5

20 15,022,278.2 –37,039,787.0 66,88,398.2 6,773,671.3 17,071,830.7

21 31,365,227.5 –24,636,261.0 6,992,516.1 7,121,654.8 17,390,125.1

22 49,158,587.6 –11,162,374.0 7,389,110.1 7,498,771.3 17,941,160.3

23 68,523,705.6 3,457,305.0 7,888,227.4 7,913,766.8 19,433,125.6

24 –30,659,758.0 –46,881,056.0 8,304,901.1 8,383,596.3 21,051,172.1

25 –14,493,133.0 –33,234,769.0 8,882,913.3 8,879,167.0 20,523,925.8

26 3,032,885.7 –18,447,471.0 9,323,873.7 9,432,300.9 21,828,511.6

27 22,052,699.9 –2,420,088.9 9,928,832.2 10,024,012.6 23,745,262.4

28 –40,777,450.0 –61,262,349.0 10,609,403.7 10,687,335.9 26,426,168.4

29 –22,017,860.0 –45,712,459.0 11,245,107.8 11,397,205.4 28,554,870.3

30 –1,696,379.3 –28,885,661.0 12,114,398.9 12,162,980.9 30,962,618.7

Fig. 4. The economic effect expressed in the change of financial result obtained as a result of implementing the scenarios

of the complex model of strategic management of heat supply enterprises, ‘000 rubles


Scenario 1 Scenario 2 Scenario 3Scenario 4 Scenario5

Time, quarters

Åconomic effect, ‘000 rubles


57

References

1. Adzhagulov E.Yu. (2006) The formation of strategy and modeling development of residential-communal management units in a small city. Doctoral thesis. Penza (in Russian).

2. Bakieva D.L. (2007) Raising the operational efficiency of enterprises in the residential-communal sphere (on the example of organizations of the Tyumen Region). Doctoral thesis. Moscow (in Russian).

3. Balandina E.V. (2009) Improving the mechanism for developing the residential-communal complex of the region. Doctoral thesis. Nizhny Novgorod (in Russian).

4. Vorobeva E.E. (2007) Budgeting in the system for managing a residential-communal operating unit of the municipality. Doctoral thesis. Moscow (in Russian).

5. Darbasov A.V. (2006) Regional peculiarities of the tariff policy in residential-communal operating units of the Republic of Sakha (Yakutia): On the example of the municipal formation “Yakutsk City”. Doctoral thesis.

Yakutsk (in Russian).

6. Makarenya T.A. (2008) Organizational and methodological aspects of regulating the market for residential-communal services: regional, municipal and micro- levels. Rostov-on-Don: SFU (in Russian).

7. Allahverdyan A.V. (2012) Managing costs of communal enterprises. Doctoral thesis. Donetsk (in Russian).

8. Brazhnikova L.N. (2010) Strategic management of the financial activity of residential-communal enterprise.

Donetsk: Southeast (in Russian).

9. Brazhnikova L.N. (2010) Strategic management of proprietary funds of residential-communal enterprises.

Donetsk: Southeast (in Russian).

10. Kulikov S.G. (2010) Improving tariff policy on the example of the regional communal enterprise for heat supply “Donetskteplokommunenergo”. Doctoral thesis. Donetsk (in Russian).

11. Myznikova M.A., Lepa R.N. (2010) Modeling optimal distribution of state investments in sub-sectors

of residential-communal enterprises. East, no 6, pp. 41–44 (in Ukrainian).

12. Lyashok Ya.A. (2015) Balanced development of branches securing the livelihood of the population in the context of forming a socially oriented economy in Ukraine. Doctoral thesis. Krasnoarmeysk (in Russian).

13. Perkova E.A. (2012) State regulation in the sphere of residential-communal management enterprises in the process of adaptation to the market environment. Doctoral thesis. Donetsk (in Ukrainian).

14. Poluyanov V.P. (2004) An organizational-economic mechanism for effective operations of enterprises in the residential-communal sector. Donetsk: Southeast (in Russian).

15. Yurchenko I.A. (2011) An organizational-economic mechanism for managing account receivables for heat supply services in enterprises of the communal heat and energy of Ukraine. Doctoral thesis. Kharkiv

(in Ukrainian).

16. Beer S. (1972) Brain of the firm: The managerial cybernetics of organization. London: Allen Lane.

17. Lyisenko Yu.G., Egorov P.V., Ovechko G.S., Timohin V.N. (2004) Economic cybernetics. Donetsk:

Southeast (in Russian).

18. Ivakhnenko A.G. (1981) An inductive method of self-organization of models for complex systems. Kiev:

Naukova Dumka (in Russian).

19. Keeney R.L., Raiffa H. (1976) Decisions with multiple objectives: Preferences and value tradeoffs. N.Y.:

John Wiley & Sons.

20. Kizim N.A., Klebanova T.S. (eds.) (2007) Adaptive models in systems of decision making. Kharkiv: INJEK

(in Russian).

21. Klebanova T.S., Guryanova L.S., Bogonikolos N., Kononov O.Yu., Bersutskiy A.Ya. (2006) Modeling the financial flows of an enterprise under conditions of uncertainty. Kharkiv: INJEK (in Russian).

22. Vovk V.M., Kizim N.A., Porokhnya V.M., Klebanova T.S., Vitlinskiy V.V., Chernyak A.I., Nebukin V.A.

(2009) Modern problems of modeling socio-economic systems. Kharkiv: INJEK (in Russian).

23. Lysenko Yu.G., Timokhin V.N., Rudenskiy R.A. (2009) Methodology of modeling viable systems in the economy. Donetsk: Southeast (in Russian).



58

24. Timokhin V.N. (2007) Methodology of modeling the economic dynamics. Donetsk: Southeast (in Russian).

25. Timokhin V.N., Lysenko Yu.G., Rudenskiy R.A. (2013) Management of strategic development of viable economic systems. Donetsk: Southeast (in Russian).

26. Forrester J.W. (1961) Industrial dynamics. Waltham (MA): Pegasus Communications.

27. Ashby W.R. (1956) An introduction to cybernetics. London: Chapman & Hall.

28. Myznikova M.A. (2013) Modeling the offer of services from residential-communal enterprises. Novelties in Economic Cybernetics, no 4, pp. 46–56 (in Russian).

29. Myznikova M.A., Kaplenko R.A. (2017) Development of a methodology for calculating the critical

maximum tariffs for residential-communal services. Manager, no 2, pp. 66–72 (in Russian).

30. Myznikova M. (2015) Constructing a model of consumer selection for services of heat supply enterprises

of the Donetsk region. Kazan Science, no 6, pp. 61–65 (in Russian).

31. Myznikova M.A. (2014) Modeling demand for enterprises of thermal communal energy // Problems

of managing the production and business activity of subjects of management. Proceedings of the VII International Scientific Conference of Young Scientists and Students of DonNTU, Donetsk, 22 May 2014,


32. Myznikova M.А. (2017) Development of models and methods for raising the effectiveness of managing heat supply enterprises in depressed territories (using the example of the Donetsk region). Doctoral thesis.

Rostov-on-Don (in Russian).

About the authors

Mariya A. Myznikova

Cand. Sci. (Econ.);

Senior Lecturer, Department of Economic Cybernetics, Donetsk National University,

24, Universitetskaya Street, Donetsk 283001, Ukraine;


Larisa N. Brazhnikova

Dr. Sci. (Econ.), Professor;

Academician, Academy of Economic Sciences of Ukraine,

2, Zhelyabova Street, Kiev 03057, Ukraine;




59

Research into the dynamics of railway track capacities in a model for organizing cargo transportation between two node stations

Nerses K. KhachatryanE-mail: [email protected]

Gayane L. BeklaryanE-mail: [email protected]

Svetlana V. BorisovaE-mail: [email protected]

Fedor A. BelousovE-mail: [email protected]

Central Economics and Mathematics Institute, Russian Academy of Sciences Address: 47, Nakhimovsky Prospect, Moscow 117418, Russia

Abstract

The article deals with a model for organizing railway transportation on a long stretch of road between two node stations connected by a large number of intermediate stations. Between two arbitrary neighboring stations, there is a railway track for temporary storage of cargo. The movement of cargo is carried out in one direction. To ensure the smooth movement of cargo, two technologies are used which are common for all stations. The fi rst technology is based on the procedure of interaction of a station with both neighboring stations and adjacent railway tracks. The second technology uses the technical capabilities of the station itself and is based on the interaction of the station with neighboring railway tracks. For cargo transportation, a simple control system is used which provides for measuring the volume of transported goods at neighboring stations with a single time lag.

This work is devoted to describing and studying the dynamics of the number of roads involved in the railway tracks. For this purpose, a system of diff erential equations is formed, the right parts of which are functions of variables describing the dynamics of the number of roads involved in the stations. The starting point for this study is previously obtained results from studying the dynamics of the number of tracks involved in the stations (a brief description of these results is given in the Introduction). What follows is the description of the dynamics of the number of roads involved in the railway tracks.



60

i – 1 i + 1i

Introduction

Transport is one of the main branches of any

state and performs the connecting, commu-

nication and supply functions. For the cor-

rect organization of traffic in the transport

network, control systems are used. Their algo-

rithms are based on mathematical models, one

of the main functions of which is the modeling

of traffic flows. A large number of publications

are devoted to mathematical modeling of traf-

fic flows. Works [1–3] describe “analog mod-

els” in which the movement of the vehicle is

similar to any physical flow (hydro- and gas-

dynamic models). There are a large number of

models designed to optimize the functioning of

transport networks [4–7]. This class of models

solves the problems of optimization of trans-

portation routes, development of optimal con-

figuration of the transport network, etc. One

of the approaches to modeling and research

of traffic flows is based on the theory of com-

petitive non-coalition equilibrium [8–11]. It

allows us to describe a fairly adequate mecha-

nism for the functioning of road networks. We

also note the approach associated with the use

of simulation and cellular automata described

in [12–15]. Recently, an alternative theory of

transport flows has been actively developed,

called the theory of three phases (classical the-

ories consider two phases: free flow and dense

flow) [16–20]. This theory can predict and

explain the empirical properties of the transi-

tion to dense flow and the resulting space-time

structures in the transport flow.

A number of publications are devoted to the

modeling of rail traffic and related transport

flows [21–27]. In particular, in works [24–27]

a model of organization of rail freight between

two node stations connected by a railway line

which contains a certain number of interme-

diate stations is investigated. It is assumed that

between stationary stations there is interex-

change railway track, where part of the cargo

can be temporarily stored (in a special storage

area). The movement of goods is carried out

in one direction. The traffic flow diagram is

shown in Figure 1.

Fig. 1. Scheme of freight traffic of railway transport

Possible variants of the dynamics (growth of the number of the roads involved on one railway tracks and falling on others) and their dependence on parameters of the model are investigated. We also study the dependence of the rate of change in the number of involved roads on the railway tracks on the model parameters. We then fi nd the parameter of control by which it is possible to provide arbitrarily small speed of growth (fall) of the number of the roads involved on all railway tracks.

Key words: station; railway track; organization of cargo transportation; mathematical model; differential

equations; dynamics; numerical realization.

Citation: Khachatryan N.K., Beklaryan G.L., Borisova S.V., Belousov F.A. (2019) Research into the

dynamics of railway track capacities in a model for organizing cargo transportation between two node

stations. Business Informatics, vol. 13, no 1, pp. 59–70.

DOI: 10.17323/1998-0663.2019.1.59.70



61

In this figure, the circles indicate the stations,

and the squares indicate the railway tracks. As

can be seen from the figure, cargo can arrive

at an arbitrary intermediate station both from

the previous station and from the railway

track, after which the cargo can be sent either

to the next station or to the railway track. Let

the number of intermediate stations be equal

to m. Denoting by 0 and m + 1 respectively the

numbers of the initial and final nodal stations,

we obtain the following set of station numbers:

{0, 1, …, m, m + 1}. Each station at any time

is characterized by the number of involved

roads. Denote by zi (t), i = 0, 1, ..., m + 1 num-

ber of roads involved in the i-th station at time

t. The maximum number of involved roads at

the stations, at which the mode of increasing

the number of roads at the expense of goods

from the railway track, is functioning, we

denote by . If the number of paths involved

exceeds the maximum value, then part of the

cargo is temporarily sent to the storage area.

The organization of cargo traffic is carried

out using two technologies.

The first technology is based on the inter-

action procedure of neighboring stations.

The following rule applies here: an arbitrary

station can send cargo to the next station if

the number of involved roads is greater than

at the next station. In this case, the inten-

sity of shipment is proportional to the dif-

ference in the number of involved roads at

these stations. Note that sending goods from

an arbitrary station (except the last) with

a certain intensity is equivalent to receiv-

ing goods with the same intensity at the next

station. Thus, each station with a number i

(1 i m) can take the cargo from the previous

station with an intensity equal to (zi – 1

– zi ),

if zi – 1

> zi and send the cargo to the next sta-

tion with an intensity equal to (zi – z

i + 1),

if zi > z

i + 1. If the first condition is violated,

the station with number i sends the cargo to

the railway track with intensity (zi – z

i – 1),

and if the second one is violated, it receives

the cargo from the railway track with intensity (z

i + 1 – z

i ). Initial node station (i = 0) takes

a cargo with intensity 1(t ) and sends it to the

next station with intensity (z 0 – z

1 ) if z

0 > z

1.

Otherwise, the initial node station additionally

takes the cargo with the intensity (z 1 – z

0 ).

Final node station (i = m + 1) accepts the

load from the previous station with intensity (z

m – z

m + 1), if z

m > z

m + 1 , and distributes it with

intensity 2(t ). If z

m < z

m + 1, then the final sta-

tion additionally distributes the cargo with

intensity (zm + 1

– zm).

The second technology is designed to use the

infrastructure capabilities of the stations and

to ensure uninterrupted movement of cargo. It

is based on the procedure of interaction of the

station with neighboring railway track located

on opposite sides of it. The second technology

for all stations, except the initial one, allows

us to increase the number of involved roads

(if it does not exceed ), and to reduce it (if it

exceeds ). The function (.), setting the speed

of change of number of involved roads within

this technology has the following properties:

on a half-line (– , 0] it is identically equal to

zero, on an interval (0, xopt

) is increasing, in

a point xopt

accepts the maximum value, on a

half-line (xopt

, + ) is decreasing, in a point

accepts zero value, and on a half-line ( , + )

is linear. For the initial node station (i = 0) the

second technology is used only for unloading.

The function 0(.), setting the speed of change

of the number of involved roads at this station

within this technology, has the following prop-

erties: on a half-line (– , ] it is identically

equal to zero, and on a half-line ( , + ) it is

linearly decreasing.

For cargo transportation, a simple control

system is used: the quantity of involved roads at

any station has to coincide with the quantity of

involved roads at the following station, with a

time log which is uniform for all stations.

Thus, the dynamics of numbers of the

involved roads at stations is set by the system of

the differential equations



62

(1)

(2)

(3)

and a control system – by nonlocal linear

restrictions

(4)

The constant will be called a characteristic

of the control system.

Definition 1. The family of absolutely con-

tinuous functions , defined on [0 , + ),

is called the solution of the traveling wave

type with characteristic (soliton solution), if

almost all functions zi (.) satisfy the

system (1) – (3) and nonlocal restrictions (4).

The class of soliton solutions is extremely

narrow. This leads to the need to properly

extend the class of soliton solutions to the class

of soliton quasi-solutions. In [24–27] two ways

of such expansion are proposed. One type of

expansion involves the assumption of discon-

tinuous soliton solutions (we call them soliton

quasi-solutions of the first type).

Definition 2. The family of absolutely contin-

uous functions , defined on [0 , + ); it

is called a soliton quasi-solution of the first type

with characteristic , if almost all t [0 , + )

functions zi (.) satisfy the system (1) – (3) and

nonlocal restrictions (4), with possible discon-

tinuities at the points

It is proved , that for any {0,

1, ... m, m + 1} system (1) – (4) with a fixed ini-

tial value at the initial time has a single

“quasi-solution” of the first type [24].

Definition 3. A soliton quasi-solution of

the first type with a characteristic is called

-quasi-solution of the first type with charac-

teristic if inequalities

,

are satisfied, for all k = 1, 2, ... .

It is proved that for any there

is -soliton quasi-solution of the first type with

a characteristic with however small > 0 [24].

The second type of expansion of soliton solu-

tions allows weakening of the control system

(implementation of nonlocal restrictions (4)

with some error). We give an exact formulation

of quasi-solutions of this type.

Definition 4. The family of absolutely con-

tinuous functions , defined on [0 , + ),

is called -soliton quasi-solution of the second

type with characteristic , if almost all

functions zi (.) satisfy the system (1) – (3) and

the condition

is satisfied.

It is proved that the solutions of the system

of differential equations (1) – (3) are limited

under the limitation of functions 1(. ) end

2(. ) [27].

In work [27] by means of computer realiza-

tion quasi-solutions of the second type were

investigated for periodic functions

1(t ) =

2(t ) = d + cos(ωt), d γ > 0,

and also functions (.) и 0(.), defined as fol-

lows:

For this purpose, the set of all solutions of

the system of differential equations (1) – (3)

was investigated. According to the results of

numerical experiments, starting from a certain

point in time > 0 the solutions of the system

(1) – (3) begin to oscillate in some neighbor-



63

hood of the value , and the components of the

solution satisfy the condition z 0 (t ) > z

1 (t ) > ...>

z m

(t ) > z

m + 1(t ) for any t [

, + ).

Moreover, there is an integer 0 < m + 1

z i (t ) > if 0 i , t [

, + ), (5)

z i (t ) < if i m + 1, t [

, + ). (6)

Numerical experiments showed that the

value depends on parameters c0, c и a, but

does not depend on parameter . Dependence

on parameters c0 and c is non-increasing: with

increasing parameter c0 value decreases to

= 0, and with increasing parameter c to

= 1. Dependence on parameters a is non-

decreasing: with its increase increases to

= m.

The dependence of the solutions of the sys-

tem of differential equations (1) – (3) on the

parameter is studied. It is shown that for an

arbitrary characteristic > 0, increasing the

parameter , it is possible to make an arbitrar-

ily small error in the performance of nonlocal

restrictions (4).

In the research carried out, it was supposed

that capacities of railway tracks (number of

the involved roads on them) are unlimited as

a result of which observation of their dynamics

was not made. However, in fact this assump-

tion is unrealistic: at least, during a long period

of time capacities of railway track have to be

limited reasonably. This work is devoted to

research into the dynamics of capacities of

railway tracks and its dependence on model

parameters.

1. Description of the dynamics of the railway tracks’ capacities

We investigated the dynamics of capacities

of railway tracks within the model described in

the Introduction. Let’s begin with their num-

bering. The railway track located between sta-

tions with numbers i and i + 1 we will designate

number i. Thus, we get the following set of rail-

way tracks numbers: {0, 1, ..., m}. The num-

ber of the involved ways on i-th railway track at

the moment of time t we will designate through

yi (t). Determine with what intensity cargo

come on the railway tracks and with what

intensity leave them. Note that the cargo can

be delivered to the railway tracks and sent from

them in the framework of both the first and the

second technology.

Within the first technology, on a stage with

number i (1 i m – 1) cargo arrives from the

station with number i with intensity (zi – z

i – 1),

if zi > z

i – 1 , and goes to the station with num-

ber i + 1 with intensity (zi + 2

– zi + 1

), if

zi + 2

> zi + 1

. On an initial railway track (i = 0)

within the first technology cargo does not

arrive. At last, on a final railway track (i = m)

within the same technology cargo arrives from

the station with number i = m with intensity

(zm – z

m – 1), if z

m > z

m – 1. The cargo is not sent

from the final railway track within this tech-

nology.

Within the second technology, on a railway

track with number i (1 i m – 1) cargo arrives

from the station with number i with inten-

sity – φ(zi

), if the number of the involved ways

at the station with number i exceeds value ,

and goes to the station with number i + 1 with

intensity φ(zi + 1

) , if the number of the involved

ways at the station with number i + 1 is less than

value (the station with number i + 1 accepts

cargo from a railway track). On an initial rail-

way track (i = 0) within the second technology

cargo arrives from the initial node station with

intensity – φ0(z

0), if the number of the involved

roads at the specified station exceeds , and

goes to the station with number i = 1 with inten-

sity φ(z1), if the number of the involved roads at

the station with number i = 1 is less . At last,

on a final railway track i = m within the second

technology cargo arrives with intensity – φ(zm)

from the station with number i = m and goes to

the final node station (i = m + 1) with intensity

φ(zm + 1

), if the number of the involved roads at

the final node station is less .



64

Thus, the dynamics of the number of the

involved roads on a railway track is described

by the following system of differential equa-

tions:

(7)

(8)

(9)

where

We investigated a system (7) – (9) on the

assumption that components of quasi-solu-

tions of the second type participate in the right

parts of the equations.

Using inequalities (5) – (6) and definitions

of functions (.) and 0(.), we will transform

the equations. In particular, from inequalities

(5) – (6) it follows that, since the moment all

composed a look (zk + 1

– zk ) sign(z

k + 1 – z

k ) in

the right part of the equations (7) – (9) will be

equal to zero. Depending on value we will

consider several cases.

The first case: = 0. It means that z 0 (t ) > ,

z i (t ) < , i = 1, ..., m + 1 for all t , and the

equation (7) – (9) take a form:

(10)

(11)

The second case: 1 < < m

(12)

(13)

(14)

(15)

The third case:

(16)

(17)

(18)

Directly from (10) – (18) it follows that in

all three cases the right parts of all equations

(except for, perhaps, equation with number )

either are positive, or are negative. Numeri-

cal experiments showed that in the first a case

( = 0) the right member of equation with

number = 0 is positive. It is connected with

the fact that this case takes place if the value

of parameter c0 is significantly more than the

value of parameter a. In the third case )

the right member of equation with number is

negative. It is connected with the fact that this

case takes place if the value of parameter c is

significantly less than value of parameter a.

Thus, in the first case the right parts of equa-

tion with number = 0 are positive, and

the right parts of other equations are nega-

tive. Therefore, in this case the number of the

involved roads on a railway track with number

= 0 will increase indefinitely, and the num-

ber of the involved roads on other railway track

will decrease indefinitely.



65

In the third case, the right part of the equa-

tion with the number is negative, and

the right parts of the remaining equations are

positive. Accordingly, in this case, the number

of involved roads on the railway track with the

number will decrease indefinitely, and

the number of involved roads on the remaining

railway tracks will increase indefinitely.

In the second case, the right parts of the

equations with numbers less than are posi-

tive, the right parts of the equations with num-

bers more than are negative. The right part

of the equation with the number can be both

positive and negative, and with certain com-

binations of parameters can be equal to zero.

Therefore, in this case only on one railway

track the number of the involved roads can-

not change over time. The number of involved

roads on the remaining railway tracks will either

increase indefinitely or decrease indefinitely.

For example, Figure 2 shows the dynamics of

the number of involved roads in the railway

tracks in case of constant functions describing

the intensity of the supply of cargo to the ini-

tial node station and the intensity of the distri-

bution of cargo from the final node station, i.e.

1(t ) =

2(t ) = d, d > 0 (case 2, equations (12) –

(15)). The number of stations is equal to 10,

respectively, the number of railway tracks is 9

(y0, y

1, ..., y

8 – the number of involved roads on

these railway tracks). The value that deter-

mines the capacity of the stations is equal to 10,

and the parameters have the following values:

= 10, a = 0.1, c0 = c = 1, d = 3. Numerous

experiments have shown that all conclusions

regarding the dynamics of the capacity of the

railway tracks, which will be given below, are

valid for any other number of stations (railway

tracks) and values .

For periodic functions 1(t ) =

2(t ) = d +

+ cos( ), d > 0 the dynamics of the num-

ber of involved roads on the railway tracks does

not change fundamentally. For example, Figure

3 shows the dynamics for the following parame-

ter values: = 10, a = 0.1, c0 = c = 1, d = 3, = 3.

50

40

30

20

10

0

-10

-20

-30

-40

-50

y1

y0

y2

y3

y4

y5

y6

y7

y8

Fig. 2. Dynamics of the number of involved roads on the railway tracks with constant functions and

50

40

30

20

10

0

-10

-20

-30

-40

-50

Fig. 3. Dynamics of the number of involved roads on the railway tracks with periodic functions and

y1

y0

y2

y3

y4

y5

y6

y7

y8

In this regard, further research will be carried

out for the case of constant and equal functions

1(t ) and

2(t ).

2. Dependence of growth rate and falling number of involved roads

on the railway tracks of the model parameters

We investigated the dependence of the growth

rate and the fall of the number of involved roads

on the railway tracks on the model parameters.

Let’s start with the parameter с0. Recall that

this parameter determines the intensity of the

shipment of cargo from the initial node station

yi , i = 0, 1, ..., 8

yi , i = 0, 1, ..., 8



66

at the zero railway track. Let (equations

(12) – (18)). As shown by numerical experi-

ments, an increase in this parameter leads to

an increase in the rate of growth of the num-

ber of involved roads on the zero railway track,

a decrease in the rate of growth of the num-

ber of involved roads on the railway tracks

with numbers 1, ..., , and an increase in the

rate of fall on the following railway tracks. At

the same time, both the decrease in the rate

of growth of the number of involved roads on

the railway tracks with numbers 1, ..., and

the increase in the rate of fall on the follow-

ing railway tracks weaken with the increase

in the number of railway tracks. This trend

can be seen in Figure 4, where the parameter

value с0 is increased to two, with unchanged

values of other parameters ( = 10, a = 0.1, c0 = 2, c = 1, d = 3).

50

40

30

20

10

0

-10

-20

-30

-40

-50

50

40

30

20

10

0

-10

-20

-30

-40

-50

Let us proceed to the study of the dependence

of the growth (fall) of the number of involved

roads on the railway tracks on the parameter c.

This parameter determines the intensity of the

shipment of cargo from any intermediate sta-

tion with the number i = 1, ..., m on the rail-

way tracks. Sending cargo to the railway track

is carried out if the number of involved roads

at the station is greater than the value , that

determines the capacity of the station. Accord-

ing to (5), this condition is satisfied at stations

with numbers i = 0, ..., . Thus, the station

with number i = 0, ..., sends cargo to the

railway track with number i = 0, ..., . Let’s

remember that the value depends on param-

eter c: with its increase the value decreases to

= 1. Therefore, a small increase in parame-

ter c which is not leading to reduction of value

leads to an increase in the growth rate of the

number of involved roads on railway tracks with

numbers i = 1, ..., and to reduction of growth

of the number of the involved roads on the rail-

way track with number i = 0. On railway tracks

with numbers i = + 1, ..., m an increase in

speed of fall of the number of involved roads

is observed. At the same time as an increase in

the growth rate of the number of involved roads

on railway tracks with numbers i = 1, ..., ,

and increase in speed of fall of the number of

y1

y0

y2

y3

y4

y5

y6

y7

y8

y1

y0

y2

y3

y4

y5

y6

y7

y8

Fig. 4. Dynamics of the number of the involved roads on railway tracks at increase of value of parameter (double increase)

Fig. 5. Dynamics of the number of involved roads on railway tracks with increase of value

of parameter (multiple increase)

Let’s remember that the value depends

on parameter с0: at its increase the value

decreases to = 0. Therefore, in the process

of increases to this parameter, growth of the

number of involved roads on all stages railway

tracks, except for initial, is replaced with fall-

ing numbers. This trend can be seen in Figure 5

(the equations (10) – (11)). In it the value of

parameter с0 is increased up to 60 at invaria-

ble values of other parameters ( = 10, a = 0.1, c0 = 60, c = 1, d = 3).

yi , i = 0, 1, ..., 8

yi , i = 0, 1, ..., 8



67

involved roads on railway tracks with numbers

i = + 1, ..., m weakens at increase in number

of a railway track.

If the increase in parameter c leads to the

reduction of value , then the following trend is

observed: in the process of an increase in param-

eter c gradually on railway tracks on which there

was an increase in the growth rate of the number

of involved roads there is a reduction of growth

rate of the number of involved roads up to fur-

ther falls, except for railway track to numbers

i = 0.1 (Figure 6). Thus, since some value of param-

eter c, on all railway tracks except zero and the

first, there is a decrease in number of the involved

roads. On zero and first railway tracks there is a

growth of number of the involved roads, and the

growth rate on the first increases (Figure 7).

In Figure 6, the value of the parameter c is

increased to two ( = 10, a = 0.1, c0 = 1, c = 2,

d = 3), and in Figure 7 – to 60, with unchanged

values of other parameters ( = 10, a = 0.1, c0 = 1, c = 60, d = 3).

Let’s pass to a research of dependence of

growth (fall) of the number of involved roads on

railway tracks from parameter a. Let’s remem-

ber that this parameter determines intensity

of receipt of cargo by the second technology

(from a railway track), and this technology is

applied if the number of involved roads at the

station are less than value . According to (6),

this condition is satisfied at stations with num-

bers i = + 1, ..., m + 1. As was stated above,

within the second technology the station with

number i + 1 accepts cargo from a railway track

with number i. Let’s remember that the value

depends on parameter a: with its increase the

value increases to . Therefore, a small

increase in parameter a, which is not lead-

ing to increase in value , leads to an increase

in speed of fall on railway tracks with numbers

i = , ..., m. On railway tracks with numbers

i = 0, ..., – 1 an increase in the growth rate of

the number of involved roads is observed. At the

same time as increase in speed of fall on railway

tracks with numbers i = , ..., m, and increase

in growth rate of number of involved roads on

the previous railway tracks weakens with the

reduction of the number of railway track. If the

increase in parameter a leads to an increase in

value , then the following trend is observed: in

the process of increases in parameter a gradually

on railway tracks on which there was an increase

in the speed of fall in the number of involved

roads there is a reduction of speed of fall in the

number of involved roads up to further growth,

except for the last railway track to number

i = m (Figure 8). Thus, since some value of

parameter a, on all railway tracks except the

last, there is a growth of number of the involved

roads. On the last railway track, there is a falling Fig. 6. Dynamics of the number of involved roads on railway tracks at increases of value of parameter (double increase)

Fig. 7. Dynamics of the number of involved roads on railway tracks at increases of value of parameter (multiple increase)

50

40

30

20

10

0

-10

-20

-30

-40

-50

50

40

30

20

10

0

-10

-20

-30

-40

-50

y1

y0

y2

y3

y4

y5

y6

y7

y8

y1

y0

y2

y3

y4

y5

y6

y7

y8

yi , i = 0, 1, ..., 8

yi , i = 0, 1, ..., 8



68

number of involved roads, and the speed of fall

increases (Figure 9).

In Figure 8, the value of parameter is increased

to 0.5 ( = 10, a = 0.5, c0 = c = 1, d = 3), and in

Figure 9 – to 10, with unchanged values of other

parameters ( = 10, a = 10, c0 = 1, c = 60, d = 3).

At last, we investigated the dependence of

growth (fall) of volume of railway tracks on

parameter . Let’s remember that change

of this parameter does not change value .

According to the research conducted in [27]

since timepoint , an increase in parameter

leads to a reduction as differences (zi – ),

i = 0, ..., and ( – zi ), i = + 1, ..., m. Thus,

an increase in parameter leads to reduction of

the growth rate on railway tracks with numbers

i = 0, ..., – 1, and to reduction of speed of

fall in the number of involved roads on railway

tracks with numbers i = + 1, ..., m. The same

impact is made by an increase in this parame-

ter and number of involved roads on a railway

track with number i = , only with the differ-

ence that the number of involved roads on this

railway tracks can both grow, and fall, or not

change. For example, the dynamics at the fol-

lowing values of parameters ( = 30, a = 0.1, c0 = c = 1, d = 3) is given in Figure 10.

Thus, an increase in parameter can reduce

both growth, and fall of the number of involved

roads on all railway tracks.

Fig. 8. Dynamics of the number of involved roads on railway tracks with increases of value of parameter (fivefold increase)

Fig. 9. Dynamics of the number of involved roads on railway tracks with increases of value of parameter (multiple increase)

Conclusion

This article is devoted to research into the

dynamics of capacities of railway tracks in a

model for the organization of cargo transporta-

tion between two node stations. Earlier in works

[24–27] the dynamics of capacities of stations

was investigated (number of involved roads on

them). In the research carried out, it was sup-

posed that capacities of railway tracks are unlim-

ited and for that reason observation of their

dynamics was not made. In this work, the system

of differential equations describing the number

of involved roads on railway tracks is presented

and investigated. As it appeared, from some

Fig. 10. Dynamics of the number of involved roads on railway tracks with increase of value of parameter

50

40

30

20

10

0

-10

-20

-30

-40

-50

50

40

30

20

10

0

-10

-20

-30

-40

-50

50

40

30

20

10

0

-10

-20

-30

-40

-50

y1

y0

y2

y3

y4

y5

y6

y7

y8

y1

y0

y2

y3

y4

y5

y6

y7

y8

y1

y0

y2

y3

y4

y5

y6

y7

y8

yi , i = 0, 1, ..., 8 y

i , i = 0, 1, ..., 8

yi , i = 0, 1, ..., 8



69

time point, the number of involved roads on

all railway tracks, except one, either increases,

or decreases. At the same time, the quantity of

railway tracks both with increasing, and with

decreasing number of involved roads depends

on a number of parameters of the model. The

dependence of the growth rate and fall of num-

ber of the involved roads on model parameters is

investigated. We revealed the parameter, which

if increased makes it possible to achieve simul-

taneous reduction of both growth rate and speed

of fall in the number of involved roads on all

railway tracks. This parameter characterizes the

intensity of interaction of the neighboring sta-

tions within the first technology of the organiza-

tion of freight traffic.

References

1. Inose H., Hamada T. (1975) Road traffic control. Tokyo: University of Tokyo Press.

2. Sukhinova A.B., Trapeznikova M.A., Chetverushkin B.N., Chubarova N.G. (2009) Two-dimensional macroscopic model of traffic flows. Mathematical Models and Computer Simulations, vol. 21, no 2, pp. 118–126 (in Russian).

3. Kholodov Ya.A., Kholodov A.S., Gasnikov A.V., Morozov I.I., Tarasov V.N. (2010) Modeling of traffic flows – Topical problems and ways to solve them. Proceedings of MIPT, vol. 2, no 4, pp. 152–162 (in Russian).

4. Galaburda V.G. (1983) Improving transportation technologies and an increase in rail capacity. Moscow: MIIT (in Russian).

5. Galaburda V.G. (1985) Optimal planning of cargo flows. Moscow: Transport (in Russian).

6. Aven O.I., Lovetsky S.E., Moiseenko G.E. (1985) Optimization of transport flows. Moscow: Nauka (in Russian).

7. Vasilieva E.M., Igudin R.V., Livshits V.N. (1987) Optimization of planning and management of transport systems. Moscow: Transport (in Russian).

8. Shvetsov V.I. (2009) Algorithms for distributing traffic flows. Automation and Remote Control, vol. 70, no 10, pp. 1728–1736.

9. Leventhal T., Nemhauser G.L., Trotter L. (1973) A column generation algorithm for optimal traffic assignment. Transportation Science, no 7, pp. 168–176.

10. Lo H.K., Chen A. (2000) Traffic equlibrium problem with route-specific costs: formulation and algorithms. Transportation Research, Part B, vol. 34, no 6, pp. 493–513.

11. Bar-Gera H. (2002) Origin-based algorithm for the traffic assignment problem. Transportation Science, vol. 36, no 4, pp. 398–417.

12. Yamada T., Russ B.F., Castro J., Taniguchi E. (2009) Designing multimodal freight transport networks: A heuristic approach and applications. Transportation Science, vol. 43, no 2, pp. 129–143.

13. Cremer M., Ludwig J. (1986) A fast simulation model for traffic flow on the basis of Boolean operations. Mathematics and Computers in Simulation, no 28, pp. 297–303.

14. Blank M.L. (2000) Exact analysis of dynamical systems arising in models of traffic flow. Russian Mathematical Surveys, vol. 55, no 3, pp. 562–563.

15. Buslaev A.P., Gasnikov A.V., Yashina M.V. (2012) Selected mathematical problems of traffic flow theory. International Journal of Computer Mathematics, vol. 89, no 3, pp. 409-432.

16. Daganzo C.F. (1997) Fundamentals of transportation and traffic operations. N.Y.: Elsevier Science.

17. Kerner B.S. (1999) Congested traffic flow: Observations and theory. Transportation Research Record, vol. 1678, pp. 160–167.

18. Kerner B.S. (1999) Theory of congested traffic flow: Self-organization without bottlenecks. Transportation and Traffic Theory. London: Elsevier Science, pp. 147–171.

19. Kerner B.S. (2009) Introduction to modern traffic flow theory and control. The long road to three-phase traffic theory. Springer.



70

20. Munoz J.C., Daganzo C.F. (2002) Moving bottlenecks – a theory grounded on experimental observation //

Transportation and Traffic Theory in the 21st Century: Proceedings of the 15th International Symposium on Transportation and Traffic Theory. Adelaide, Australia, 16–18 July 2002. P. 441–461.

21. Rubtsov A.O., Tarasov A.S. (2009) Modeling of railway transportation in Russia. Proceedings of the Institute for Systems Analysis of the Russian Academy of Sciences, no 46. pp. 274–278 (in Russian).

22. Gainanov D.N., Konygin A.V., Rasskazova V.A. (2016) Modelling railway freight traffic using the methods

of graph theory and combinatorial optimization. Automation and Remote Control, vol. 77, no 11,

pp. 1928–1943.

23. Levin D.Yu. (2010) Modeling of transportation processes. World of Transport and Transportation, vol. 8, no 5,


24. Beklaryan L.A., Khachatryan N.K. (2006) Traveling wave type solutions in dynamic transport models.

Functional Differential Equations, vol. 13, no 12, pp. 125–155.

25. Beklaryan L.A., Khachatryan N.K. (2013) On one class of dynamic transportation models. Computational Mathematics and Mathematical Physics, vol. 53, no 10, pp. 1466–1482.

26. Khachatryan N.K, Akopov A.S. (2017) Model for organizing cargo transportation with an initial station

of departure and a final station of cargo distribution. Business Informatics, no 1, pp. 25–35.

27. Khachatryan N.K, Akopov A.S., Belousov F.A. (2018) About quasi-solutions of traveling wave type in models

for organizing cargo transportation. Business Informatics, no 1, pp. 61–70.

About the authors

Nerses K. Khachatryan

Cand. Sci. (Phys.-Math.);

Leading Researcher, Laboratory of Dynamic Models of Economy and Optimization,


47, Nakhimovsky Prospect, Moscow 117418, Russia;


Gayane L. Beklaryan

Cand. Sci. (Econ.);

Senior Researcher, Laboratory of Computer Modeling of Social and Economic Processes;




Svetlana V. Borisova

Cand. Sci. (Phys.-Math.);

Senior Researcher, Laboratory of Dynamic Models of Economy and Optimization,




Fedor A. Belousov

Researcher, Laboratory of Dynamic Models of Economy and Optimization,






71

INFORMATION SECURITY

Analysis and forecast of undesirable cloud services traffic

Marina V. TumbinskayaE-mail: [email protected]

Bulat I. BayanovE-mail: [email protected]

Ruslan Zh. RakhimovE-mail: [email protected]

Nikita V. KormiltcevE-mail: [email protected]

Alexander D. UvarovE-mail: [email protected]

Kazan National Research Technical University named after A.N. TupolevAddress: 10, Karl Marx Street, Kazan 420111 Russia

Abstract

These days one of the main problems that must be solved to ensure information security in cloud services for corporations as well as for individual clients is to correctly identify and predict hacking in the network traffi c. This paper presents statistics on information security threats, provides classifi cation of information security threats for cloud services, identifi es hackers’ goals, and proposes countermeasures.

A vital task is to develop an eff ective method that could be used to protect cloud services from various network threats, as well as to analyze the network traffi c. For these purposes, we chose a method based on an additive time series model, which allows us to predict the undesirable network traffi c. To test this method, we obtained quantitative parameters for the undesirable traffi c by simulating a network attack and collecting empirical data that describe this process. We used special software that simulates a network attack, and software that records and processes all the empirical data needed for the research.

Using the data obtained, we analyzed the effi ciency of the method based on the additive time series model. We demonstrated that this method is also applicable for research into the general dynamics of the number of network attacks in cyberspace. This method also allows us to reveal how the dynamics of the number of hacker network attacks depends on season, date, or time. The results show that, based on data describing the network traffi c, one can identify and predict the undesirable hacker threats.


72

Introduction

Development of the infrastructure

of modern enterprises causes an

increased demand for cloud tech-

nologies, because they are convenient, reason-

ably priced, mobile, quick, and reliable. Cloud

technologies allow us to use cloud services.

A cloud service [1] is an Internet service that

makes it possible for its clients to outsource

the maintenance of some elements of IT infra-

structure [2].

According to the RightScale statistics, 95%

of organizations used one or another cloud ser-

vice model in 2017 [3]. According to Orange

Business Services experts, the market for

cloud services in Russia comes to about 24.6

Bn. roubles [4]. It was shown [2, 5] that mod-

ern IT companies are uncompetitive if they

fail to use cloud technologies, thus foregoing

profits. Cloud services have long been used in

large corporations (Google Disk, iCloud from

Apple, Cloud mail.ru).

Cloud services make it necessary to solve

information security problems, since new tech-

nologies lead to the emergence of a large num-

ber of threats and vulnerabilities in information

security systems. According to a Kaspersky

Laboratory poll [6], 13% of Russian organiza-

tions face issues related to cloud infrastructure

security at least once a year. Out of those com-

panies, 32% lost their data due to such inci-

dents. Therefore, it is crucial to ensure infor-

mation security in cloud services.

The proposed novel method for analyzing

and predicting the network traffic based on the


additive time series model and integrated into

security tools can ensure the necessary security

level for regular storage, thus protecting it from

various network attacks. This constitutes the

scientific novelty of the paper. Unfortunately,

many existing data security methods cannot

reliably predict undesirable network traffic.

1. Possibility of interpreting

the proposed method in WAF

As shown in [7], the majority of hacker

attacks are based on typical hacker methods,

which are brought to perfection. Therefore, we

need to develop methods that employ continu-

ous learning, and such methods should gradu-

ally replace the signature analysis. It was also

noted [7] that some developers of web appli-

cation firewalls (WAF) focus on renewing the

signatures rather than on the signature analy-

sis. To create a security model that ensures the

necessary security level, WAF needs an exten-

sive database of the undesirable traffic signa-

tures and actions that can be applied to all types

of web applications. The proposed method for

analyzing and predicting the network traffic

based on the additive time series model can

be integrated into complicated WAFs in the

future. Here the main goal will not be to pre-

dict the hacker’s and legitimate user’s actions,

but to create a security model based on the

URL as well as on the parameters and cookies.

After the security model is developed, it needs

to be tested, i.e., the traffic should be analyzed

to prevent a hacker’s exploiting both known

and unknown vulnerabilities.

Key words: forecasting; DDOS attack; cloud services; network traffic; modeling; additive time

series model; autocorrelation function; error estimation.

Citation: Tumbinskaya M.V., Bayanov B.I., Rakhimov R.Zh., Kormiltcev N.V., Uvarov A.D.

(2019) Analysis and forecast of undesirable cloud services traffic. Business Informatics, vol. 13,

no 1, pp. 71–81

DOI: 10.17323/1998-0663.2019.1.71.81


73

2. Classification of security threats

to cloud services

Let us consider the classification of security

threats to cloud services. Table 1 presents the

most common threats according to [6]. Pos-

sible hackers’ goals and security measures are

presented for each threat. No single method

alone can prevent all types of threats; there-

fore, it is impossible to block the threats com-

pletely. Statistical data for each threat that was

successful can be stored in the system and used

for future analysis and development of new

security systems.

3. Simulating a network attack

To analyze the network traffic coming into

the network nodes, information security spe-

cialists install ad hoc software at the network

nodes. In this research, Wireshark software

(v.2.6.1) was used. This software allows us to

capture and analyze the network traffic for the

most common network protocols (TCP, UDP,

HTTP, etc.).


Table 1. Types of security threats to cloud services,

hackers’ goals, and security measures

# Security threat to cloud services Hackers’ goal Security measures

1. Data theft Accessing a database (e.g., e-mail addresses of users)

Database decentralization and data encryption with an SSL certificate

2. Data loss Database modification or erasing information Data backup, restricted access

3. Account theft / hacked services

Database modification or erasing information Two-factor authentication

4. Unprotected nterfaces and API Complete access to the database Authentication, access control,

encryption

5. DDOS attacks Preventing authorized users from accessing the cloud service Access control

6. Undesirable insider Database access Access control

7. Cloud services used by hackers

Access to the cloud computing resources

Restriction of the system’s computing power

Papers [8, 9] contain the network traffic data

that describe DDOS attacks. However, there is

not enough data there for the purposes of this

research; therefore, we collected the necessary

data by simulating a network attack based on

the algorithms presented in [10]. We used two

nodes of a configured network. One of them

was used as the victim’s device, and the other

one as the hacker’s device. Virtual machines

installed on the same computer served as those

devices. Wireshark was installed on the victim’s

virtual machine, and LOIC (an open-source

code for DDOS attacks)1, which creates unde-

sirable traffic, was installed on the hacker’s vir-

tual machine.

In our research, we assumed that hackers

attack the network multiple times with vari-

ous initial configurations of the malware, and

we did not rule out the possibility that the vic-

tim could access the network as well. The net-

work stream (the number of network packets

per second) through the victim’s network node

is presented in Figure 1.

1 https://www.darknet.org.uk/2017/10/loic-download-low-orbit-ion-cannon-ddos-booter/


74

Number of packets, unit



Fig. 1. Number of network packets passing through the victim’s node, per second

Fig. 2. Number of network packets from the hacker’s node, per second

1 15 29 43 57 71 85 99 113

127

141

155

169

183

197

211

225

239

252

267

281

295

309

323

337

351

365

379

393

5000

4000

3000

2000

1000

0

1 29 57 85 113

141

169

197

225

253

281

309

337

365

393

421

449

477

505

533

561

589

617

645

673

701

729

757

785

As we can see from Figure 1, we cannot dis-

tinguish the stream from a particular user from

the overall network stream; therefore it is rec-

ommended to consider network streams from

particular users.

For convenience of analysis, it is possible to

filter the network stream through the victim’s

node and separate the packets coming from the

hacker’s node. The network stream from the

hacker’s node is presented in Figure 2. It repre-

sents a physical process, the intensity of which

periodically increases by several orders of mag-

nitude and describes the actions of a particu-

lar user.

We can find out the address of the node of the

hacker attacking the network by analyzing the

parameters describing the incoming network

traffic, for example, the density of the distri-

bution of the number of packets by their size in

bits. Table 2 presents the data obtained by sim-

ulating network attacks and desirable network

traffic using Wireshark software.

The simulation results show that, when the

network traffic is undesirable, more than 92%

of the packets are 40–79 bits in size. At the

same time, when the traffic is desirable, the

percentage of the packets of this size is about

39%, while more than 42% of the packets have

the size between 1280–2559 bits, and about

11% are sized between 640–1279 bits. It is also

suspicious when the traffic is extremely inten-

sive (in terms of the number of packets per unit

of time) or it has other untypical parameters.

As a sample for analysis, we chose the number

of network packets coming from the hacker’s

node per second (Figure 2).

350030002500200015001000500

0


75


4. Predicting the network attacks

with time series analysis

For statistical analysis, we chose the method

based on time series analysis. According to the

Cisco annual report on cybersecurity [11], in

year 2018, 39% of organizations used auto-

mated tools to prevent hacker attacks, and the

rest of them used machine learning (artificial

intelligence) [12–15].

We solved the problem of predicting the net-

work attacks using a time series additive model.

This model assumes that each level of the time

series (F ) can be presented as a sum of three

components: a trend (T ), a seasonal compo-

nent (S ), and a random component (E ):

F = T + S + E. (1)

To determine the trend component, linear

regression was used:

у = a x + b, (2)

where y – the trend value;

x – the lag;

a and b – the regression coefficients.

In Equation (2), coefficients a and b are

determined from the previous values in the

original sample using the following equation:

(3)

a = y – b · x, (4)

where – the mean lag value;

– the mean value in the original sample.

Figure 3 presents the original data on the

number of network packets coming from the

hacker’s node alongside the trend line obtained

by Equation (2), where a = 0.973, b = 615.87.

The trend line goes up because of the increase

in the intensity of the network traffic.

Now we have to determine the seasonal

component, which is periodical and can be

obtained from the autocorrelation function

(ACF). Figure 4 presents a plot of the auto-

Table 2. The percentage of received packets by packet size for desirable traffic and during a network attack

# Packet size Percentage of packets received by packet size for desirable traffic

Percentage of packets received by packet size during

a network attack

1. 0–19 0.00% 0.00%

2. 20–39 0.00% 0.00%

3. 40–79 39.06% 92.79%

4. 80–159 3.81% 0.48%

5. 160–319 0.93% 3.30%

6. 320–639 1.35% 3.25%

7. 640–1279 11.05% 0.16%

8. 1280–2559 42.90% 0.02%

9. 2560–5119 0.82% 0.00%

10. 5120 and more 0.08% 0.00%


76


correlation function of. the lag number. The

dashed line corresponds to the white noise (the

boundary of the statistical significance of cor-

relation coefficients is the error of the autocor-

relation function). The autocorrelation func-

tion was calculated for a time interval of up to

30 lags.

Analysis of the autocorrelation function

shows that the original data is periodical. There

is a high correlation for lags 22 and 23. There-

fore, for the seasonal component in the addi-

tive model the period will be about 23 lags.

Thus, the length of one season is N = 23 (the

lag’s number can take values n = 1, 2, ..., N),

where one lag corresponds to one second.

The values of seasonal component Sn are deter-

mined as mean values of the differences between

current the value Fn and the trend component T

n

calculated for each lag number n:

(5)

where k – the season number;

K – the total number of seasons.

Then the total number of lags for the entire

time series is M = N K.

Using the values obtained for the trend com-

ponent (2) and the seasonal component (5), we

can calculate the predicted values for F using

Equation (1) (in this model, the random com-

ponent is not considered) [16]. Figure 5 presents

the plots for sample values Fn and predicted val-

ues F. The discrepancies between the plots for F

and Fn can be evaluated by calculating the mean

absolute percentage error (MAPE).

Value of ACF

Fig. 3. The number of network packets received from the hacker’s node per second, including the trend component

Fig. 4. Autocorrelation function with the account of the white noise

1 18 35 52 69 86 103

120

137

154

171

188

205

222

139

256

273

290

307

324

341

358

1,0000,8000,6000,4000,2000,000

-0,200-0,200-0,600

1 4 7 10 13 16 19 22 25 28

Time, sec.


350030002500200015001000500

0

Experiment Trend

Time lag, units


77


This estimate cannot be used to calcu-

late the error of the prediction model used in

this research because the current test sample

includes values close to 1. Therefore, we used

the root-mean-square error (RMSE) instead;

it is equal to 353. This follows from the follow-

ing equation:

(6)

where N – the original sample size;

y – the predicted value,

– the current value.

The value obtained indicates that the pre-

diction model is less than optimal. To make

Fig. 5. Current values for the test sample and predicted values for the number of network packets per second

Fig. 6. The number of network packets coming from the hacker’s node per second, with the account of the trend component for two closest seasons from the original sample

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47Time, sec.

1600140012001000800600400200

0

our additive predicting model more accurate,

we used the trend and seasonal components

from earlier seasons (the most recent ones), as

shown in Table 3 and Figure 6.

Both the amplitude and the duration of these

seasons are close to the corresponding val-

ues for the following season, and this fact can

improve the quality of the prediction.

Figure 7 presents the plots for future actual

values for the test sample and the predicted

values, taking into account the corrections

to the components of the time series additive

model.

In this case, the estimated RMSE is 201,

which is a considerable improvement com-

pared to the earlier RSME of 353. This leads


0 5 10 15 20 25

Forecast Experiment

Time, sec.


2000

1500

1000

500

0

Experiment Trend


78

us to the conclusion that a predictive model of

network attacks is much more accurate when

it is based on recent experimental data rather

than on the entire sample.

To calculate the relative error for the pre-

diction model, we can calculate the ratio of

the RMSE estimate to the maximum value in

the test sample. Here we chose the maximum

instead of the mean value because the sam-

ple considered contains many values close to

1. This leads to a relatively small mean value,

which makes it impossible to estimate the rela-

tive error (MAPE) reliably. We found that the

ratio of the RMSE to the maximum value in

the test sample is 13%.

Therefore, the proposed prediction model

of undesirable network traffic has a reasonably

small relative error, and it can serve as an effi-

cient tool for the detection of network attacks.

If necessary, the proposed model for the pre-

Fig. 7. Actual values for the test sample and predicted values for the number of network packets per second, with corrections

16001400120010008006004002000

diction of DDOS attacks could be used to study

the general dynamics of the number of DDOS

attacks in cyberspace [17]. If we use the num-

ber of attempted DDOS attacks in each quar-

ter of years 2017 and 2018 as empirical train-

ing data, we can predict the number of DDOS

attacks in the first half of year 2019.

The analysis of the data presented in Fig-

ure 8 shows that there are two periods in the

dynamics of the number of DDOS attacks,

namely 60 and 7 days. Apparently, the activity

peaks (Feb 15, 2019; April 10, 2019, and June

5, 2019) of the envelope curve fall between rel-

atively long holidays (March, May, and June).

Short-scale periodic peaks are probably caused

by the activity during particular days of the

week. Therefore, a relatively simple prediction

model allows us to find a connection between

the periods in DDOS attacks and the calendar

features for 2019.

Fig. 8. Predicted numbers of DDOS attacks

01.01.2019 31.01.2019 02.03.2019 01.04.2019 01.05.2019 31.05.2019 30.06.2019

140012001000800600400200

0Time, days

Number of DDOS attacks, unit

0 5 10 15 20 25

Forecast Experiment

Time, sec.




79

Table 3. Estimation of the trend and seasonal components

# Current values, number of packets per second

Trend component estimate, number of packets per second

Seasonal component estimate, number of packets per second

1. 65 661 –5962. 21 662 –6413. 9 663 –6544. 18 663 –6455. 1088 664 4246. 1398 665 7337. 1301 666 6358. 1363 667 6969. 1343 668 675

10. 1375 669 70611. 1283 670 61312. 1378 671 70713. 1387 672 71514. 1304 673 63115. 1276 674 60216. 1302 675 62717. 1295 676 61918. 1380 677 70319. 1391 678 71320. 1062 679 38321. 15 679 –66422. 23 680 –65723. 11 681 –67024. 10 682 –67225. 19 683 –66426. 24 684 –66027. 13 685 –67228. 36 686 –65029. 36 687 –65130. 1313 688 62531. 1342 689 65332. 1360 690 67033. 1439 691 74834. 1380 692 68835. 1290 693 59736. 1384 694 69037. 1329 695 63438. 1306 695 61139. 1315 696 61940. 1296 697 59941. 1309 698 61142. 1298 699 59943. 93 700 –60744. 37 701 –66445. 21 702 –68146. 9 703 –694



80


Also, we should note that efficiency of the

proposed model is higher when DDOS attacks

have almost identical statistical parameters. If

each implementation of a DDOS attack differs

statistically, it is harder to detect and predict

the hacker’s actions.

Conclusion

This paper reports the results of network traf-

fic analysis aimed at predicting the threats in

cloud services. The statistics on information

security threats to data storage and transmis-

sion that we present here validate the need for

the development of new methods of data pro-

tection. Such methods typically use ad hoc

hardware and software to analyze the informa-

tion security threats. We implemented the mal-

ware that simulated network attacks, as well as

the software that captured and processed the

empirical data we needed for this study. We sim-

ulated a network attack (a DDOS attack) and

saved the necessary parameters to files conven-

ient for analysis and further processing. Out of

many prediction models, we chose the additive

time series model. The results obtained with

the help of this model show that if we know

the behavior of the statistical parameters of

different implementations of a DDOS attack,

we can detect and predict the hacker’s actions

for this type of attacks. The high efficiency of

the proposed model is proven by comparison

of the predicted values with the future actual

values. The model’s accuracy is characterized

by the RMS error, which is equal to 201. The

results of our research demonstrate that statis-

tical methods of network traffic analysis can be

employed in the tools used to protect the cloud

services from various network attacks.

References

1. Maksimov K.V. (2018) Planning of activities in the IT-company in conditions of uncertainty taking into

account the use of cloud services Applied Informatics, vol. 13, no 1, pp. 25–31 (in Russian).

2. Tumbinskaya M.V. (2017) Providing protection from targeted information in social networks. Mordovia University Bulletin, vol. 27, no 2, pp. 264–288 (in Russian).

3. Weins K. (2017) Cloud computing trends: 2017 state of the cloud survey. Available at:

https://www.rightscale.com/blog/cloud-industry-insights/cloud-computing-trends-2017-state-cloud-

survey (accessed 10 October 2018).

4. Krupin A. (2014) We go into the clouds: Russian premiere of the IaaS-platform Flexible Computing Express. Available at: https://servernews.ru/813983?k292300 (accessed 10 October 2018) (in Russian).

5. Tumbinskaya M.V. (2017) Process of distribution of undesirable information in social networks. Business Informatics, no 3, pp. 65–76.

6. Tadviser (2018) Security threats in the cloud. Available at: http://tadviser.ru/a/170054 (accessed

10 October 2018) (in Russian).

7. Baranov P.A., Beybutov E.R. (2015) Securing information resources using web application firewalls.

Business Informatics, no 4, pp. 71–78.

8. Garcia S., Grill M., Stiborek J., Zunino A. (2014) An empirical comparison of botnet detectionmethods.

Computers and Security, no 45, pp. 100–123.

9. Kosenko M.Yu., Melnikov A.V. (2016) Issues of protecting business information systems from botnets

attacks. Voprosy Kiberbezopasnosti, no 4, pp. 20–28 (in Russian).

10. Gneushev V.A., Kravets A.G., Kozunova S.S., Babenko A.A. (2017) Modeling network attacks

of attackers in a corporate information system. Industrial Automatic Control Systems and Controllers, no 6,


11. Cisco (2018) Cisco annual report on cybersecurity for 2018. Available at: https://www.cisco.com/c/ru_ru/

about/press/press-releases/2018/03-12.html (accessed 10 October 2018) (in Russian).


81


About the authors

Marina V. Tumbinskaya

Cand. Sci. (Tech.);

Associate Professor, Department of Information Protection Systems,

Kazan National Research Technical University named after A.N. Tupolev,

10, Karl Marx Street, Kazan 420111, Russia;


Bulat I. Bayanov

Student, Kazan National Research Technical University named after A.N. Tupolev,



Ruslan Zh. Rakhimov




Nikita V. Kormiltcev




Alexander D. Uvarov




12. Glushenko S.A. (2017) An adaptive neuro-fuzzy inference system for assessment of risks to an

organization’s information security. Business Informatics, no 1, pp. 68–77.

13. Kartiev S.B., Kureichik V.M. (2016) Classification algorithm based on random forest principles for

forecasting problem. Software & Systems (Programmnye produkty i sistemy), no 2, pp. 11–15 (in Russian).

14. Afanas’ev A.P., Dzyuba S.M., Emelyanova I.I. (2017) Horner’s Scheme for investigation of solutions

of differential equations with polynomial right-hand side. Business Informatics, no 2, pp. 33–39.

15. Tomilin A., Tumbinskaya M., Tregubov V., Smolevitskaya M. (2017) The BESM-6 virtualization project.

Proceedings of the 2017 Fourth International Conference on Computer Technology in Russia and in the Former Soviet Union (SoRuCom 2017). Moscow, 3–5 October 2017, pp. 241–245.

16. Pevtsova T.A., Ryabukhina E.A., Gushchina O.A. (2015) Calculation of seasonality index. Mordovia University Bulletin, vol. 25, no 4, pp. 18–36 (in Russian).

17. Kupreev O., Badovskaya E., Gutnikov A. (2018) DDoS attacks in the third quarter 2018. Available at:

https://securelist.ru/ddos-report-in-q3-2018/92512/ (accessed 12 November 2018) (in Russian).


The IEEE Conference Series on Business Informatics is the leading international forum for state-of-the-art research in Business Informatics. The 21st IEEE CBI 2019, held in huge, old and interesting city Moscow, calls for submissions in the multidisciplinary fi eld of Business Informatics, and welcomes a multitude of theoretical and practical perspectives and mind-sets on today’s challenges of the digital transformation.

21st IEEE International Conference on Business Informatics (IEEE CBI 2019) July 15–17, 2019, Moscow, Russia

http://www.cbi2019.moscow/ http://www.cbi-series.org

Call for Papers

The IEEE CBI series encourages a broad understanding of Business Informatics research, and intends to further its many diff erent facets, theoretical foundations and experiential body of knowledge. In doing so, the CBI series has proven to be a fertile ground for research with high impact, and a hub for multidisciplinary research with contributions from Management Science, Organization Science, Economics, Information Systems, Computer Science, and Informatics.

In line with this, the CBI series aims to provide a forum for researchers and practitioners from various fi elds and backgrounds that contribute to the construction, use and maintenance of information systems and their organizational use contexts. A dedicated goal of the CBI series of conferences is to bring together researchers from diff erent fi elds and disciplines to stimulate discussion, synergy and collaboration. Accordingly, CBI conferences use a format that enables in depth discussions among researchers during the conference. The benefi ts of such a crossdisciplinary conception are contrasted by a challenge: Authors who submit a paper take the risk to be assessed by standards that are diff erent from those they are used to in their own research communities. The conference organisation accounts for this challenge by involving renowned track chairs and programme committee members from diff erent fi elds and disciplines.

List of Tracks

• Enterprise Modelling, Engineering and Architecture

• Business Process Management

• Information Systems Engineering• Artifi cial Intelligence for Business• Information Management

• Business Analytics and Business Data Engineering

• Industry 4.0 (Industry Applications)• Business Innovations and Digital Transformation• Data-Driven Business Applications• General Topics

Submission

https://edas.info/newPaper.php?c=25431

Manuscripts must be in English and are restricted to 10 pages in IEEE 2-coloumn template (A4).

In the submission form, select the conference track that best fi ts with your manuscript’s topic, or choose “General Topics” if not covered by the tracks.

Review process is double-blind; your names and affi liations must not be explicitly listed in the manuscript. However, you should keep all cited references in the submission’s bibliography intact.

Further information for submission:

https://cbi2019.moscow/ Accepted conference content will be submitted for inclusion into IEEE Xplore as well as other Abstracting and Indexing (A&I) databases. Selected papers will be invited to submit an extended version to a special issue of the EMISA Journal (subject to separate review).

Important Dates

Full paper submission March 1, 2019

Notifi cation of paper acceptance May 10, 2019

Camera-ready paper submission June 1, 2019

Articles should be topical and original, should outline tasks

(issues), describe key results of the author’s research and

appropriate conclusions.

Manuscripts are submitted via e-mail: [email protected].

MANUSCRIPT REQUIREMENTS

TEXT FILES should be submitted in electronic form, as a MS

Word document (version 2003 or higher).

LENGTH. Articles should be between 20 and 25 thousand

characters (incl. spaces).

FONT, SPACING, MARGINS. The text should be in Times

New Roman 12 pt, 1.5 spaced, fit to the width, margins: left –

25 mm, all other – 15 mm.

TITLE of the article should be submitted in native language

and English.

AUTHORS’ DETAILS are presented in native language and

English. The details include:

Full name of each author

Position, rank, academic degree of each author

Affiliation of each author, at the time the research was

completed

Full postal address of each affiliation (incl. postcode / ZIP) E-mail address of each author.

ABSTRACT are presented in native language and English.

The abstract should be between 200 and 300 words.

The abstract should be informative (no general words),

original, relevant (reflects your paper’s key content and research

findings); structured (follows the logics of results’ presentation

in the paper)

The recommended structure: purpose (mandatory), design

/ methodology / approach (mandatory), findings (mandatory),

research limitations / implications (if applicable), practical

implications (if applicable), originality / value (mandatory).

It is appropriate to describe the research methods/

methodology if they are original or of interest for this particular research. For papers concerned with experimental work the data sources and data procession technique should be described.

The results should be described as precisely and informatively as possible. Include your key theoretical and experimental results, factual information, revealed interconnections and patterns. Give special priority in the abstract to new results and long-term impact data, important discoveries and verified findings that contradict previous theories as well as data that you think have practical value.

Conclusions may be associated with recommendations, estimates, suggestions, hypotheses described in the paper.

Information contained in the title should not be duplicated in the abstract. Authors should try to avoid unnecessary introductory phrases (e.g. «the author of the paper considers…»).

Authors should use the language typical of research and technical documents to compile your abstract and avoid complex grammatical constructions.

The text of the abstract should include key words of the paper.

KEYWORDS are presented in native language and English. The number of key words / words combinations are from 6 to 10 (separated by semicolons).

FORMULAE should be prepared using Math Type or MS Equation tool.

FIGURES should be of high quality, black and white, legible and numbered consecutively with Arabic numerals. All figures (charts, diagrams, etc.) should be submitted in electronic form (photo images – in TIF, PSD or JPEG formats, minimum resolution 300 dpi). Appropriate references in the text are required.

REFERENCES should be presented in Harvard style and carefully checked for completeness, accuracy and consistency.

The publication is free of charge.

AUTHORS GUIDELINES