corridor capacity analysis with mesoscopic simulation

Corridor capacity analysis with mesoscopic simulation: Erzincanprovince sample

AHMET OGUZ DEMIRIZ1,* , OSMAN UNSAL BAYRAK2 and HALIM FERIT BAYATA1

1Department of Civil Engineering, Faculty of Engineering, Erzincan Binali Yildirim University, Erzincan,

Turkey2Department of Civil Engineering, Faculty of Engineering, Ataturk University, Erzurum, Turkey

e-mail: [email protected]

MS received 9 June 2020; revised 7 October 2020; accepted 1 November 2020

Abstract. Today, the existing traffic networks cannot cope-up with the increasing transportation demands and

traffic volume and fail to meet the demands. This causes economic and social losses. Increasing transportation

demands and traffic volume in Erzincan city center of Turkey cause the main arteries and junctions to be insufficient.

In such problems, solutions are generally produced using parameters such as trip and delay durations and exhaust

emissions. Another point to be considered is to ensure the improved intersections towork as a whole. In this study, the

main arteries and the junctions on these in Erzincan city center were discussed. In this context, totally 14 intersections

were tried to be improved. Suggestions were offered regarding delay time, travel time, queue lengths, NOx and CO2

emissions, and average speed values as the design criteria. Simulations were made in AIMSUN program for the

current situation and four different scenarios. Analytical Hierarchy Process (AHP), which is one of the multi-criteria

decision-making methods for the network was employed to determine which scenario was more appropriate. At the

end of the study, decrease of up to 23% for travel time, 47% for delays, 48% for queue lengths, 11% forNOx, and 13%

for CO2 and increase of up to 30% in average speed were obtained. According to AHP, Scenario 4 had the highest

priority value among all alternatives. Moreover, the reliability of the results was ensured by supporting the meso-

scopic simulation with AHP which is a multi-criteria decision-making method.

Keywords. Intersection; traffic management; mesoscopic simulation; road network; AIMSUN.

1. Introduction

Transportation is one of the most important factors

embedded with any stages in people’s life. Transportation

demands have increased day by day along with factors such

as population growth, growth of cities and increase at

industrialization. As a result of the increasing transportation

demands, densities appear in transportation networks, and

congestion starts locally. As traffic density increases, the

parameters that negatively affect human life such as fuel

consumption, travel time, and risk of accidents increase.

The road section where the most traffic accident occurs in

our country is the intersections. According to the traffic

accident data announced by the General Directorate of

Highways, there were 186.532 accidents involving death or

personal injuries [1].

Intersections are one of the factors that mostly affect the

risk of accident, travel time, fuel consumption, and road

performance. For these reasons, along with the increasing

traffic density, an intersection is one of the factors that

needs to be improved first. Traffic simulations are used

when designing intersections. In addition to being safe and

economical, simulation is a very effective instrument for

model production and evaluation of produced models [2–5].

The researchers who have recently carried out studies on

transportation have benefited from mesoscopic simulations

to compare large-scale traffic networks and assessed their

performances [5–7]. Various mesoscopic simulation pro-

grams such as AIMSUN, VISUM, SATURN, TRACKS,

DYNAMEQ, DYNAMIT, etc. offer the opportunity of

evaluating and comparing traffic networks [7–10] (traffic

modelling guidelines). In this study, AIMSUN mesoscopic

model was used to assess and compare the current situation

and alternative scenarios of Erzincan central traffic net-

work. AIMSUN mesoscopic model is ideal for simulating

and evaluating large-scale traffic networks. Although

alternative scenarios are analyzed and assessed with AIM-

SUN, the researcher decides which alternative is the opti-

mum solution. Because simulation programs cannot

determine the parameters such as security or benefit/cost.

In engineering applications, some criteria are determined

to assess alternatives, and the alternatives are compared

with each other in terms of these determined criteria. Multi-

criteria decision-making methods are used when comparing*For correspondence

Sådhanå (2021) 46:10 � Indian Academy of Sciences

https://doi.org/10.1007/s12046-020-01533-9Sadhana(0123456789().,-volV)FT3](0123456789().,-volV)

alternatives in terms of pre-determined criteria. The most

frequently used multiple-criteria decision-making methods

are Analytic Hierarchy Process (AHP), Analytic Network

Process (ANP), Electre, Promethee, and Technique for the

Order of Prioritization by Similarity to Ideal Solution

(TOPSIS). The Analytical Hierarchy Process (AHP)

developed by Saaty [11] has become one of the most

effective multi-criteria evaluation methods that have been

used in a wide range of practical applications for various

decision-making processes. Since its introduction, the AHP

has shown a broad impact in the field of multi-criteria or

multi-attribute decision making and has been recognized

with its numerous applications such as choices, prioritiza-

tion/evaluation, resource allocation, benchmarking, and

quality management [12]. In this study, AHP was used to

determine the optimum option among the alternatives.

One of the most important issues to design effective road

networks is the junction type selection. For example, if a

junction where traffic regulatory signs can be used is

designed as signalized, the values such as delay time, travel

time, and fuel consumption increase [4]. Or if a grade-

separated junction is to be replaced by a modern junction

with high-level service, the level of service does not change

and only the cost of construction increases. Thus, an ana-

lytical approach that considers assessing multi-criteria is

essential as well as the cost-benefit analysis that assesses

the consequences of different intersection design types in

monetary terms [13].

In recent years, several studies have been carried out on

simulation applications for traffic control and management

[6, 14–16]. Some researchers have calibrated mesoscopic

simulation parameters according to their study area

[5, 7, 8, 10, 17–20]. Some others have used AHP to dis-

cover the optimum option among the alternatives they have

created [13, 14, 21–23]. However, it was observed that very

few of these studies were large-scale, and therefore, most of

them were not assessed as a whole using mesoscopic sim-

ulation. Therefore, in this article, all the corridors in a city

were investigated as a whole, the scenarios with different

types of junctions were created, and the optimum option

was revealed using the AHP method. For these purposes,

the 14 intersections with the highest density in Erzincan

city center and the link roads of these intersections were

considered as a corridor. The total corridor had a length of

about 25 km and a heavy traffic volume during working

hours. In order to increase corridor performance, the current

situation and 4 scenarios (Scenario 1: Current Situation;

Scenario 2: Optimization with Webster method, Scenario 3:

Modern Roundabout Applications; Scenario 4: Optimiza-

tion with Webster method along with roundabout applica-

tion and signalized roundabout application; Scenario 5:

Grade Separated Junction Application) were assessed and 4

alternative scenarios were created to increase the perfor-

mance. During the decision-making process, the criteria of

delay time, travel time, queue length, NOx emission, CO2

emission, speed and cost of construction were taken into

consideration. All criteria except from the cost of con-

struction were determined using the AIMSUN program.

Then, AHP was used to determine the optimum scenario

considering all the criteria.

2. Material and method

2.1 Case study- an Erzincan city

Fourteen junctions with the highest traffic density and on

the main arteries were discussed as the study areas. These

junctions were Bahcesaray, Isikpinar, Milli Egemenlik,

Esentepe, Mengucek Gazi, Adnan Menderes, Cumartesi,

Ataturk, Halit Pasa, Yildiz, Izzet Pasa, Menguceli, Terz-

ibaba, and Sumer. The whole network was regarded as a

single corridor (figure 1). In the study area, there were 13

signalized and 1 unsigned junction (figure 2).

It was observed that traffic density and traffic jams

occurred at the current intersections especially in the

morning and evening hours. At least one of these inter-

sections has been used for any transportation demands in

Erzincan. There are an average of 6 junctions within the

city bus (public transport) routes. Each of the junctions

used in the study had a different importance and cannot be

ignored. In addition to the aforementioned reasons, the

corridors that were selected covered the entire Erzincan city

center, the importance of the corridor increased with any

increasing transportation demands.

2.2 Data collection

Data collection and data analysis are very important for

traffic engineers in terms of traffic control and manage-

ment. Various preliminary investigations such as traffic

volume counts in the corridor, signaling times of intersec-

tions, and travel time were carried out during peak hours of

morning, lunchtime, and evening. In this study, the sig-

naling phase times were obtained from Erzincan Munici-

pality’s Signaling unit. Vehicle counts were made manually

for some intersections according to vehicle types (bus and

Figure 1. Test area.

10 of 12 Sådhanå (2021) 46:10

automobile), and via video camera recordings for the oth-

ers. The vehicles were counted on Monday, Wednesday,

Friday and Sunday between 7:45 and 8:45 a.m., 12:00

and1:00 p.m. 4:45 and 5:45 p.m. In-vehicle counts, the

highest traffic volume was determined between 7:45 and

8:45 a.m. on Monday. The data were analyzed and used as

input data for AIMSUN’s mesoscopic simulation software.

The traffic composition showed that 94% of total vehicles

were private cars, and 6% were public transport vehicles

(buses). Private vehicles and public transport buses were

counted separately according to their routes [24]. For

example, the traffic volumes of the Yildiz junction were

presented in figure 3.

Here, it was observed that there were 448 cars and 49

buses coming from the East-West direction (indicated as 1

numbered road) and going straight (to 4 numbered road),

170 cars and 18 buses turning to the left (to 6 numbered

road), 28 cars making a U-turn (to 8 numbered road) 52

cars and 6 buses turning to the right (to 2 numbered road). It

was also observed that there were 462 cars and 43 buses

coming from West-East direction (from 5 numbered road)

and going straight (to 8 numbered road), 243 cars and 23

buses turning to the left (to 2 numbered road), 66 cars

turning to the right (to 6 numbered road), and 31 cars

making a U-turn (to 4 numbered road). From the North-

South direction (from 3 numbered road) to the left (to 4

numbered road) 71 cars 28 buses, to the straight (to 6

numbered road) 313 cars and 20 buses, to the right (to 8

numbered road) 186 cars, U-turn. There were 43 cars that

made (to 2 numbered road). Also, there were 71 cars and 28

buses coming from the North-South direction (from 3

numbered road) turning to the left (to 4 numbered road), to

313 cars and 20 buses going straight (to 6 numbered road),

186 cars turning to the right (to 8 numbered road), and 43

cars making U-turn (to 2 numbered road). Moreover, it was

also observed that there were 123 cars and 4 buses coming

from South-North direction (from 7 numbered road) and

turning to the left (to 4 numbered road), 229 cars and 20

buses going straight (to 2 numbered road), 68 cars and 16

buses turning to the right (to 8 numbered road) and 8 cars

making U-turn (to 6 numbered road).

In figure 4, the phase durations of the signaling system at

the Terzibaba junction were presented. At Terzibaba junc-

tion, the duration lasted for 62 s, the yellow light time

between red and green light was 2 s, and protection red (all

red) was 2 s. The starting point of the arrows shown in red

indicated the point where the traffic lights were located, and

the direction of the arrow indicated the permissible transi-

tion phases.

Different speed limits are applied in Erzincan city center

as 50 km/h, 60 km/h, and 70 km/h. The speed limits

administered in the traffic network were presented in fig-

ure 5. Lane using is reduced in some streets as parking is

permitted in many arteries of the city. For example, in

figure 6, the sections of the lanes in the arteries connected

to the safety junction that cannot be used due to parking

were marked in red. In simulation modeling to be created,

the current situation will be reflected as it is regarding the

speed limits and parking conditions of the arteries in the

traffic network.

2.3 Model building and calibration

The image of Erzincan City Center taken using the Google

Earth program was transferred to the AIMSUN software on

Figure 2. The intersections in the test area (Reference: Googles

Maps).

Sådhanå (2021) 46:10 of 12 10

a scale, and the accuracy of the road lengths was measured

by car. The collected data were entered into the software,

and the current situation was modeled. The models created

with simulation software are needed to be calibrated to

reflect the real situation. The default parameter values are

changed till the error between actual and simulated mea-

sures such as flow or travel time less than the required

threshold value [3, 25]. However, in current traffic

engineering, no accepted standards exist to determine when

a model can be considered to be suitably validated or cal-

ibrated [26]. Different methods such as GEH formula

[15, 27–29], relative error [25, 30], etc. used for calibration.

In this study, traffic volumes were used for the calibration

parameter. Traffic volumes were counted according to their

routes (figure 3) and simulation values were obtained from

data collection points. Counted values and simulated values

were assessed according to the GEH formula in Table 1.

2.4 Webster method

The Webster method is a method used to regulate the phase

duration and period of the signals at junctions controlled by

signalization and is also known as the British method.

Signalization optimization was provided in several studies

using Webster method [31–33]. Through the Webster

method, signal optimization was separately applied to each

junction with signalization, and the most successful among

them were used in this study. These junctions were Ataturk,

Halit Pasa, Menguceli, Cumartesi, Terzibaba and Izzet

Pasa. The new phases created with the Webster method for

one of these intersections was presented in figure 7. In

figure 7, the phase durations and rights of ways in the

phases were shown in details.

Figure 3. Yildiz intersection, traffic volumes of directions [24].

Figure 4. Terzibaba intersection phase durations [24].

10 of 12 Sådhanå (2021) 46:10

In figure 7, the new phases of Ataturk Junction created

with the Webster method were presented. In the upper part

of the figure, the permitted transitions in Signal 1 and

Signal 2 phases were shown. Signal 1 and Signal 2 phase

diagrams were given in the middle part, and traffic light

durations in the phases were presented in the lower part.

According to figure; Signal 1 had 26-second red light

duration, 4-second yellow light duration, and 46-second

green light duration. Signal 2 had 50-second red light

duration, 3-second yellow light duration and 23-second

green light duration.

Since the physical conditions of the Yıldız Junction

which locates at the midpoint of the entire road network

and has the highest traffic density are insufficient, green

wave application could not be applied. Hence, Webster

signal optimization has been done instead of the green

wave.

Figure 5. Speed limits of road sections in the traffic network.

Figure 6. Lanes to be parked in Ataturk junction.

Table 1. GEH values.

Yildiz Junction

GEH

WEST-EAST EAST-WEST NORTH-SOUTH SOUTH-NORTH

TOTALCar Bus Car Bus Car Bus Car Bus

Observed values 802 66 698 73 613 48 428 40 2768

Simulation values 806 62 684 98 599 44 386 41 2720

GEH value 0.14 0.50 0.53 2.70 0.57 0.59 2.08 0.16 0.92

Criteria GEHT\ 5 YES YES YES YES YES YES YES YES YES

Sådhanå (2021) 46:10 of 12 10

2.5 AHP method

In many engineering applications, the final decision is

based upon the assessment of a number of alternatives in

terms of a set of criteria. Multi-criteria decision-making

methods help decision-makers to find solutions that best

corresponds to their goals. Different methods have been

developed to solve multi-criteria analysis problems. One of

these developed methods is the AHP method

[12–14, 21, 23, 34]. The Analytic Hierarchy Process (AHP)

is a multi-criteria decision-making approach developed by

Saaty [11, 35, 36], and used for solving the complex

decision problems [13, 34, 37]. Many researchers have

come to the final conclusion using AHP in their studies

[14, 23, 38, 39].

The criteria determined to be used for the comparison

and assessment of the scenarios were given below. The

definitions of the criteria were given as used for mesoscopic

simulation network results in AIMSUN software.

1. Delay Time: The delay time in the software is the

average delay time of vehicles per kilometer. The

average of the delay times for all vehicles is calculated

and then converted to time per kilometer. The unit is

second/kilometer.

2. Travel time: It is the ratio of the total travel time of all

vehicles from the beginning of the simulation to the total

trip distance, and is expressed in second/km.

3. Average Queue Length: It is the average queue length

that occurs in the whole traffic network during the

simulation in the program and is expressed in vehicles.

4. NOx Emission: Nitrogen Oxide emission is the amount

of NOx emitted by all vehicles leaving the traffic

network in kilometers and expressed in gram/kilometer

in the software.

5. CO2 Emission: Carbon Monoxide emission is the

amount of CO2 emitted by all vehicles leaving the

traffic network in kilometers and expressed in gram/

kilometer in the software.

6. Speed: It is the average speed of all vehicles leaving the

traffic network in AIMSUN Software and expressed in

kilometer/hour.

7. Cost of Construction: The costs of the regulations to be

created at junctions are as important as the benefits. For

this reason, the construction cost was included as a

criterion since the benefit/cost ratio should also be

regarded while making regulations.

In order to weigh the determined criteria, decision-

makers compared the criteria with the paired comparison

method. In the comparison process, 1-9 scale was used.

After creating a paired comparison matrix, the criteria

weights were calculated. The paired comparison matrix was

presented in Table 2.

In Table 2, indicating the paired comparison matrix, the

superiority of the criteria in the rows against the criteria in

the column were given at the junction points. For example,

when the Table was analyzed, it was possible to notice that

the travel time and cost of construction were slightly more

significant than the delay time, and the other criteria were

less significant. The weights calculated as result of the

paired comparison matrix were given in the last column of

the Table. According to the calculated weights, the most

significant criterion was the cost of construction and the

least significant one was the queue length.

3. Alternative preference

Junctions are road sections where traffic flow coming from

more than one direction intersects, joins and separates.

Planning is made regarding the factors of comfort, safety

and capacity. Here, the safety factor means preventing a

vehicle from colliding with a pedestrian or a vehicle

coming from a different direction and eliminating the ele-

ments that may create an accident hazard. The comfort

factor means having no situations that may require sudden

braking or acceleration and minimizing the time lost

between braking and acceleration. And capacity is a desired

situation for junction planning.

The junctions are grouped under two main types as at-

grade junctions and grade-separated junctions. After the

traffic volume in junctions exceeds a certain level, the

capacity of the at-grade junctions becomes insufficient.

This can increase delay times and queue lengths at junc-

tions as well as causing serious traffic accidents. Grade

separated junctions are constructed in cases when at-grade

junctions remain insufficient. At-grade junctions are the

ones that appear as a result of the intersection of traffic

flows from different directions on the same plane. The

majority of traffic accidents or problems occur in at-grade

junctions. For this reason, much attention should be paid to

its design. At-grade junctions are grouped under 4 as

Figure 7. Ataturk intersection, the phases created with the

Webster method.

10 of 12 Sådhanå (2021) 46:10

uncontrolled, signalized, modern roundabouts, and signal-

ized roundabouts. Uncontrolled junctions are the ones

where there is no control mechanism. Traffic flow depends

on factors such as the behavior, knowledge, and initiative of

the drivers. The signalized junctions are the ones where

traffic movements are controlled with traffic signals. In

these junction types, various methods such as the Webster

method are used to calculate signal times to prevent

excessive delay, queue length, and fuel consumption.

Modern roundabouts, on the other hand, are the ones where

traffic flows regularly and continuously around a traffic

island without signalization and with specific rules with

traffic regulatory signs. Signalized roundabouts do not have

enough space on the traffic island and have signalization to

evacuate the traffic on the island and prevent traffic jams to

appear.

In the light of the above information, 5 scenarios together

with the current situation were regarded for the junctions in

Erzincan city center. These are:

• Current situation of all junctions (Scenario 1).

• Signal optimization of some junctions with Webster

method (Scenario 2).

• Design of some junctions as modern roundabouts

(Scenario 3).

• Design of some junctions as modern roundabouts,

signaling optimization of some junctions with Webster

method, and signalization of a junction adding a traffic

island (Scenario 4).

• Design of some junctions as the subway (Scenario 5).

The details of the scenarios were presented in Table3.

3.1 Modern roundabout applications

In the current traffic network, all junctions that provided

necessary geometrical conditions and had high traffic

density were modeled as modern roundabouts. The junc-

tions converted into modern roundabouts were Ataturk,

Yildiz, Bahcesaray, Isikpinar, Izzet Pasa, and Menguceli.

Created modern roundabouts were separately presented in

the figures, and the maximum possible radius was used in

all. The roundabout with a radius of 50 m was modelled at

Table 2. Paired comparison matrix.

Criteria

Delay

time

Travel

time

Queue

length

NOx

emission

CO2

emission Speed

Cost of

construction Weight

Delay time 1 1/2 5 4 4 2 1/2 0.186

Travel time 2 1 3 5 4 2 1/3 0.212

Queue length 1/5 1/3 1 1/2 1/2 1/2 1/4 0.047

NOx emission 1/4 1/5 2 1 1/2 1/3 1/6 0.048

CO2 emission 1/4 1/4 2 2 1 1/3 1/4 0.064

Speed 1/2 1/2 2 3 3 1 1/3 0.116

Cost of construction 2 3 4 6 4 3 1 0.327

Lambda (max):

7.396

CI: 0.066 RI: 1.32 CR: 0.05 (\ 0.10)

Table 3. The scenarios to be created.

Junctions Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5

Bahcesaray Current Situation Current Situation Roundabout Roundabout Current Situation

Isikpinar Current Situation Current Situation Roundabout Roundabout Current Situation

Milli Egemenlik Current Situation Webster Method Current Situation Webster Method Current Situation

Mengucek Gazi Current Situation Current Situation Current Situation Signalized Roundabout Current Situation

Adnan Menderes Current Situation Current Situation Current Situation Current Situation Current Situation

Cumartesi Current Situation Webster Method Current Situation Webster Method Current Situation

Ataturk Current Situation Webster Method Roundabout Roundabout Grade Separated

Halit Pasa Current Situation Webster Method Current Situation Webster Method Grade Separated

Yildiz Current Situation Current Situation Roundabout Roundabout Grade Separated

Izzet Pasa Current Situation Current Situation Roundabout Roundabout Grade Separated

Menguceli Current Situation Webster Method Roundabout Roundabout Current Situation

Terzibaba Current Situation Webster Method Current Situation Webster Method Current Situation

Sumer Current Situation Current Situation Current Situation Current Situation Current Situation

Sådhanå (2021) 46:10 of 12 10

Ataturk, 20 m at Yildiz, Bahcesaray, and Isikpinar junc-

tions, and 26 m at Izzet Pasa and Menguceli junctions were

modeled.

In figure 8, an example of a modern roundabout model

for the Ataturk junction was presented. The flow inside the

roundabout was prioritized, and a ‘‘give way’’ regulatory

sign was added to the arteries.

3.2 Signalized roundabout applications

It was observed that the waiting times in the Hospital

junction with four legs four-phase signaling are higher

rather than other junctions. Webster’s method was tried at

the junctions that needed improvement, but no satisfactory

results were obtained. It was also observed that the existing

geometry was not sufficient for making modern round-

abouts. For these reasons, the signaling system with 4

phases was reduced to 2 phases adding a traffic island to the

junction. The new signaling with 2 phases was created

using Webster’s method. There was no signalization in the

designed traffic island and there was a ‘‘give way’’ traffic

regulatory sign.

As shown in figure 9, the two opposed legs on Signal 1

operated in the same phase and two other legs in Signal 2

operated on the same phase. Traffic lights were located at

the beginning of the red arrows. The red arrows indicated

permitted transitions. Signal 1 had a 27-second red light


green light duration. Signal 2 had a 33-second red light

Figure 8. Modern roundabout application in Ataturk junction.

Figure 9. Mengucek Gazi signalized roundabout phase

durations.

Figure 10. Ataturk grade-separated junction application.

60

70

80

90

100

110

120

130

140

Scenari

o 1

Scenari

o 2

Scenari

o 3

Scenari

o 4

Scenari

o 5

Ave

rage

Que

ue L

engt

h

3600

3700

3800

3900

4000

4100

4200

Scenari

o 1

Scenari

o 2

Scenari

o 3

Scenari

o 4

Scenari

o 5

NO

x (G

ram

) (a)

(b)

Figure 11. a Delay time. b Travel tim.

10 of 12 Sådhanå (2021) 46:10


green light duration.

3.3 Grade separated junction application

In Turkey, if the volume of traffic in a junction is more than

1800 vehicle/hour/lane, grade-separated junctions should

be considered (46). For that reason, in this study, the sub-

way application was carried out in the east-west direction

where the traffic volume was the highest in Ataturk, Halit

Pasa, Yildiz, and Izzet Pasa junctions in Erzincan city

center. The junction models created were presented in

figure 10.

In figure 10, the 3-dimensional view of the Ataturk

junction was presented on the left and the view from the top

was presented on the right. As the existing geometry at

Ataturk junction permitted, two-lane subways designing. A

single-lane subway was designed according to the existing

geometry at Halit Pasa junction. At Yildiz junction, a sin-

gle-lane subway was designed according to the existing

geometry. At Izzet Pasa junction, a two-lane subway was

designed according to the current geometry.

4. Results

For current conditions and alternative scenarios, delay time,

travel time, queue length, CO2 and NOx emissions, and

speed were obtained and assessed using AIMSUN meso-

scopic simulation. Subsequently, AHP was used as a multi-

criteria decision-making method to determine the appro-

priate solution scenario.

4.1 Assessment of mesoscopic simulation outputs

Scenarios were modeled separately in AIMSUN software,

and the results were obtained in accordance with pre-de-

termined criteria. The scenarios created were revealed

comparatively in terms of the criteria. The graphics of delay

time, travel time, queue length, NOx emission, CO2 emis-

sion, and speed criteria were separately presented below.

According to the values presented in the delay time

chart in figure 11a and b, Scenario 4 had the lowest delay

value, and Scenario 1 was indicated the case with the

highest delay time. According to the values shown in the

10

15

20

25

30

Scenari

o 1

Scenari

o 2

Scenari

o 3

Scenari

o 4

Scenari

o 5

Del

ay T

ime

(Sec

ond)

707580859095

100

Scenari

o 1

Scenari

o 2

Scenari

o 3

Scenari

o 4

Scenari

o 5

Trav

el T

ime

(Sec

ond)

(a)

(b)

Figure 12. a Queue length. b NOx emission.

3384000

3484000

3584000

3684000

3784000

3884000

Scenari

o 1

Scenari

o 2

Scenari

o 3

Scenari

o 4

Scenari

o 5

CO

2 (G

ram

)

3537394143454749515355

Scenari

o 1

Scenari

o 2

Scenari

o 3

Scenari

o 4

Scenari

o 5

Spee

d (K

m/h

)

(a)

(b)

Figure 13. a CO2 emission. b speed.

Sådhanå (2021) 46:10 of 12 10

travel time graph, Scenario 4 had the lowest travel time

value, and Scenario 1 had the highest.

According to the values presented in the queue length

chart in figure 12a and b, it seemed that the highest queuing

value appeared in Scenario 1, and the lowest queuing

appeared in Scenario 4. According to the values shown in

the NOx emission graph, it was noticed that the maximum

NOx emission appeared in Scenario 1 and the lowest in

Scenario 4.

According to the values shown in the CO2 emission

graph in figure 13a, it was noticed that the highest CO2

emission appeared in Scenario 1, and the lowest in Scenario

4. According to the values presented in the speed graph in

figure 13b, Scenario 1 was the one with the lowest average

speed and Scenario 5 had the highest average speed.

4.2 Assessment of AHP results

In order to weigh the pre-determined criteria, decision-

makers compared the criteria with the paired comparison

method. In the comparison process, 1-9 scale was used. The

criteria weights were calculated after creating a paired

comparison matrix. The combined weight of all criteria, in

other words, the result matrix, was given in Table 4.

The combined weight of all criteria, in other words the

result matrix, was given in Table 4. According to this

Table, the weights of the criteria as delay time, travel time,

NOx emission, and cost of construction were noticed to be

highest in Scenario 4. Moreover, the highest weights in

terms of CO2 emission and queue length criteria were

obtained in Scenario 4 and Scenario 5. When the AHP

method was employed, it was revealed that Scenario 4 with

the highest combined weight was the optimum option.

Besides, Scenario 1, in other words, the current situation,

was revealed as the most inappropriate scenario.

5. Conclusion

The purpose of this study was to investigate all main

arteries and the junctions of Bahcesaray, Isikpinar, Milli

Egemenlik, Esentepe, Mengucek Gazi, Adnan Menderes,

Cumartesi, Ataturk, Halit Pasa, Yildiz, Izzet Pasa, Men-

guceli, Terzibaba, and Sumer intersections in Erzincan city

center and to offer suggestions to improve delay time,

travel time, queue length, NOx emission, CO2 emission,

and speed values. In accordance with this purpose, the

current situation of the corridors in Erzincan city center and

4 different alternative scenarios were offered and investi-

gated. According to the APS method, the most suit-

able junction type related to the parameters given above

were determined among the alternatives. The results

obtained from the study are presented below.

• In simulations, ‘‘Scenario 1’’ decreased delay time to

9%, ‘‘Scenario 3’’ to 34%, ‘‘Scenario 4’’ to 47%, and

‘‘Scenario 5’’ to 41%.

• In simulations, ‘‘Scenario 2’’ decreased travel time to



• In simulations, ‘‘Scenario 2’’ decreased queue length to



• In simulations, ‘‘Scenario 2’’ decreased NOx emission

to 5%, ‘‘Scenario 3’’ to 9%, ‘‘Scenario 4’’ to 11%, and


• In simulations, ‘‘Scenario 2’’ decreased CO2 emission

to 6%, ‘‘Scenario 3’’ to 10%, ‘‘Scenario 4’’ and


• In simulations, ‘‘Scenario 2’’ increased average speed

to 15%, ‘‘Scenario 3’’ to 26%, ‘‘Scenario 4’’ to 30%,

and ‘‘Scenario 5’’ to 31%.

• The most appropriate alternative according to the AHP

method was obtained in Scenario 4. The improvements

in Scenario 4 were reconstructed for Bahcesaray,

Isikpinar, Ataturk, Yildiz, Izzet Pasa, and Menguceli

junctions as modern roundabouts, optimization of Milli

Egemenlik, Cumartesi, Halit Pasa, and Terzibaba

junctions with the Webster method, and construction

of a signalized roundabout to Mengucek Gazi junction.

• The employability of the AIMSUN software meso-

scopic simulation model was noticed to be possible in

corridor analysis.

In this study, the number and routes of pedestrians on the

corridors and public transportation demands were not

Table 4. Combined weight of all criteria.

Criteria Criteria weights Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5

Delay time 0.186 0.044 0.062 0.213 0.351 0.330

Travel time 0.212 0.043 0.095 0.203 0.343 0.316

Queue length 0.047 0.042 0.062 0.199 0.349 0.349

NOx emission 0.048 0.041 0.102 0.219 0.398 0.240

CO2 emission 0.064 0.037 0.093 0.203 0.333 0.333

Speed 0.116 0.037 0.099 0.207 0.319 0.338

Cost of construction 0.327 0.381 0.350 0.147 0.085 0.037

Combined weights 0.152 0.171 0.188 0.263 0.225

10 of 12 Sådhanå (2021) 46:10

modeled. Further research is possible to be carried out

including passenger demands, stops, and passenger routes

for public transportation. Also, in this study, only the peak

hours on days without precipitation were regarded. On

snowy and rainy days, capacity changes at junctions and

roads as a result of the changing characteristic features such

as reducing drivers’ acceleration-deceleration and increas-

ing the following distance. For this reason, it is also rec-

ommended to carry out further researches for peak hours on

rainy days.

References

[1] GDH (General Directorate of Highways) (2018) TrafikKazalari Ozeti. GDH, Ankara, Turkey (in Turkish)

[2] Park B and Schneeberger J 2003Microscopic simulation

model calibration and validation: case study of VISSIM

simulation model for a coordinated actuated Signal system.

Transp. Res. Rec. J. Transp. Res. Board. 1856, 185–192[3] Siddharth S and Ramadurai G 2013 Calibration of VISSIM

for Indian heterogeneous traffic conditions. Procedia-Soc.Behav. Sci. 104: 380–389

[4] Bayrak O U and Abret N E 2019 A multi-criteria decision for

determining the appropriate junction design type: AHP

approach with microsimulation. Feb-Fresenius Environ.Bull. 7183

[5] Jiang Z and Huang Y X 2009. Parametric calibration of

speed–density relationships in mesoscopic traffic simulator

with data mining. Inf. Sci. 179(12): 2002–2013[6] Zlatkovic M, Zlatkovic S, Sullivan T, Bjornstad J and

Shahandashti S K F 2019. Assessment of effects of street

connectivity on traffic performance and sustainability within

communities and neighborhoods through traffic simulation.

Sustain. Cities Soc. 46: 101409[7] Toledo T, Cats O, Burghout W and Koutsopoulos H N 2010.

Mesoscopic simulation for transit operations. Transp. Res.Part C Emerg. Technol. 18(6): 896–908

[8] Jamshidnejad A, Papamichail I, Papageorgiou M and De

Schutter B 2017. A mesoscopic integrated urban traffic flow-

emission model. Transp. Res. Part C Emerg. Technol., 75:45–83

[9] Celikoglu H B and Dell’Orco M 2007. Mesoscopic simula-

tion of a dynamic link loading process. Transp. Res. Part CEmerg. Technol. 15(5): 329–344

[10] Dell’Orco M, Marinelli M and Silgu M A 2016. Bee colony

optimization for innovative travel time estimation, based on

a mesoscopic traffic assignment model. Transp. Res. Part CEmerg. Technol., 66: 48–60

[11] Saaty T L 1977 A scaling method for priorities in hierar-

chical structures. J. Math. Psychol. 15(3): 234–281[12] Vaidya OS and Kumar S 2006 Analytic hierarchy process:

An overview of applications. Eur. J. Oper. Res. 169(1): 1–29[13] Murat Y S, Arslan T, Cakici Z and Akcam C 2016.

Analytical hierarchy process (AHP) based decision support

system for urban intersections in transportation planning. In:

Using Decision Support Systems for Transportation Plan-ning Efficiency (pp. 203–222). IGI Global

[14] Rahimov K, Motamadnia A and Samadi S 2016. Technical

and economic evaluation of Pinavia interchange in compar-

ison with roundabout intersection by AIMSUN. Civ. Eng. J.2(3): 102–112

[15] Bayata H F and Bayrak O U Yeni Yapılması Planlanan bir

Kavsagın Mikro-Simulasyon ile Degerlendirilmesi. Erzincan

Universitesi Fen Bilimleri Enstitusu Dergisi, 11(3): 550–559[16] Zhang R and Yao E 2019. Mesoscopic model framework for

estimating electric vehicles’ energy consumption. Sustain.Cities Soc. 47: 101478

[17] Xu Y, Song X, Weng Z and Tan G 2014. An entry time-

based supply framework (ETSF) for mesoscopic traffic

simulations. Simul. Modell. Pract. Theory 47: 182–195

[18] Bennett J and Betts B 2017, August). Aimsun saturation flow

calibration for hybrid simulation: challenges and recommen-

dations. In: Australian Institute of Traffic Planning andManagement (AITPM) National Conference, 2017, Mel-bourne, Victoria, Australia

[19] Ciuffo B, Casas J, Montanino M, Perarnau J and Punzo V

2013. Gaussian process metamodels for sensitivity analysis

of traffic simulation models: Case study of AIMSUN

mesoscopic model. Transp. Res. Rec. 2390(1): 87–98[20] Schenk M, Tolujew J and Reggelin T2009. Mesoscopic

modeling and simulation of logistics networks. IFAC Proc.Vol. 42(4): 582–587

[21] Xu T, Moon D H and Baek S G 2012. A simulation study

integrated with analytic hierarchy process (AHP) in an

automotive manufacturing system. Simulation, 88(4):

450–463

[22] Alp S and Engin T Trafik Kazalarının Nedenleri ve SonuclarıArasındaki Iliskinin Topsis ve AHP Yontemleri KullanılarakAnalizi ve Degerlendirilmesi. Istanbul Ticaret Universitesi

Fen Bilimleri Dergisi, 10(19), 63

[23] Baric D, Pilko H and Strujic J (2016). An analytic hierarchy

process model to evaluate road section design. Transport,

31(3), 312–321

[24] Demiriz A O 2019 Coridor Capacity Analysis with Meso-scopic Simulation: Erzincan Province Sample, Ataturk

University and Graduate School of Natural and Applied

Sciences, Master Thesis, Erzurum (in Turkish)

[25] Yu L, Yu L, Chen X and Guo J 2006 Calibration of VISSIM

for bus rapid transit systems in Beijing using GPS data. J.Public Transp.. 9(3): 13

[26] Hellinga B R 1998 Requirements for the calibration of traffic

simulation models. Proc. Can. Soc. Civ. Eng. 4:

211–222

[27] Dalaff C, Ebendt R, Erdmann J, Gurczik G and Touko L C

2013 Benchmarking SUMO Generated Traffic Simulation

Results Based on GEH Statistic. In: 1st SUMO UserConference (p. 54)

[28] Alomari A H, Al-Deek H, Sandt A, Rogers J H and Hussain

O 2016 Regional evaluation of bus rapid transit with and

without transit signal priority. Transp. Res. Rec. J. Transp.Res. Board. 2554, 46–59

[29] Karakikes I, Spangler M and Margreiter M 2016 Motorway

network simulation using bluetooth data. Transp. Telecom-mun. J. 17(3): 242–251

[30] Song G, Yu L and Zhang Y 2012 Applicability of traffic

microsimulation models in vehicle emissions estimates: Case

study of VISSIM. Transp. Res. Rec. J. Transp. Res. Board.2270, 132–141

Sådhanå (2021) 46:10 of 12 10

[31] Wu Y, Chen H and Zhu F 2019. DCL-AIM: Decentralized

coordination learning of autonomous intersection manage-

ment for connected and automated vehicles. Transp. Res.Part C Emerg. Technol., 103, 246–260

[32] Ceylan H, Baskan O, Ceylan H and Haldenbilen S 2011.

Yaklasık Hesaplama Metodu ile Sinyalize KavsaklardaGecikme Bilesenlerinin Matematiksel Cozumu. Pamukkale

Universitesi Muhendislik Bilimleri Dergisi, 13(2): 279–288

[33] Murat Y S 2006 Sinyalize kavsaklardaki tasıt gecikmelerinin

bulanık mantık ile modellenmesi. IMO Teknik Dergi3903(3916), 258

[34] Triantaphyllou E and Mann S H 1995 Using the analytic

hierarchy process for decision making in engineering applica-

tions: some challenges. Int. J. Ind. Eng. Appl. Pract. 2(1): 35–44[35] Saaty T 1980 The Analytic Hierarchy Process (New York:

McGrawHill, 1980). MATH Google Scholar

[36] Saaty T L 1994 Fundamentals of decision making andpriority theory with the analytic hierarchy process. Vol. VI.Universitas Pittsburgh. USA

[37] Yu J, Wang L and Gong X 2013 Study on the status

evaluation of urban road intersections traffic congestion base

on AHP-TOPSIS modal. Procedia-Soc. Behav. Sci. 96:

609–616

[38] Effat H A and Hassan O A 2013 Designing and evaluation of

three alternatives highway routes using the Analytical

Hierarchy Process and the least-cost path analysis, applica-

tion in Sinai Peninsula, Egypt. Egypt. J. Remote Sens. SpaceSci. 16(2): 141–151

[39] Abdi, M R and Labib A W 2003 A design strategy for

reconfigurable manufacturing systems (RMSs) using analyt-

ical hierarchical process (AHP): a case study. Int. J. Prod.Res. 41(10): 2273–2299

10 of 12 Sådhanå (2021) 46:10

corridor capacity analysis with mesoscopic simulation

Documents