quantitative techniques for decision making in … · quantitative technique for decision making in...

QUANTITATIVE TECHNIQUES

FOR DECISION MAKING

IN CONSTRUCTION

S.L. Tang Irtishad U. Ahmad

Syed M. Ahmed Ming Lu

# m *. * & m, * t H O N G KON G UNIVERSIT Y PRES S

Hong Kong University Pres s 14/F Hing Wai Centre 7 Tin Wan Praya Road

Aberdeen Hong Kong

© Hong Kong University Press 2004

ISBN 962 209 705 7

All rights reserved. No part of this publication may be reproduced o r transmitted, in any form o r by any means,

electronic or mechanical, including photocopy, recording , or any information storag e or retrieval system, without prio r

permission in writing from th e publisher .

British Library Cataloguing-in-Publication Dat a A catalogue record for thi s book is available from th e British Library

Secure On-lin e Orderin g h ttp ://www. hkupress. org

Printed and bound by Condor Production Co . Ltd., Hong Kong, China

CONTENTS

Preface

Chapter 1 :

Chapter 2:

Chapter 3:

Chapter 4:

Chapter 5:

Chapter 6 :

Chapter 7 :

Chapter 8 :

Chapter 9 :

Chapter 10 :

Chapter 11:

Chapter 12 :

Analytic Hierarchy Process I

Analytic Hierarchy Process II

Decision Theory using EMV Criterion

Decision Theory using EUV Criterion

The Value of Information i n Decision Models

Inventory Modelling I

Inventory Modelling II

Dynamic Programmin g

Simulation I

Simulation II

Information System s and Process of Decision Making

Planning and Scheduling Decisions

vii

1

19

35

49

65

73

85

101

125

143

151

163

I V i CONTENT S

Bibliography 18 1

Answers to Selected Exercise Questions 18 5

About the Authors 21 9

/ \

46 HIERARC]^Y|

ROCESS 1 /K

1.1 Wha t I s Analytic Hierarch y Process (AHP) ?

The "Analytic Hierarchy Process" (or AHP in short), a mathematical tool for management decisio n making , was introduced b y Thomas L . Saaty (197 7 and 1980). The mathematical technique is capable of handling a large number of decision factor s an d provides a systematic procedur e o f ranking man y decision variables. It is a decision analysis technique which can be very useful in construction management. This chapter will firstly give a brief description of the theory of AHP. Cases will then be used to illustrate how this analysis technique can be applied in the field o f construction .

1.2 Mathematica l Theory of AHP

1.2.1 W e will use the selection o f tenders as an example in explaining th e AHP theory. Suppos e tha t ther e ar e n factor s i n considerin g whethe r a tende r should b e accepte d o r not . Thes e n factor s hav e differen t importanc e contributing t o the acceptance or unacceptance o f the tender . In assessin g the importance o f each factor , pairwis e comparison s ar e used s o tha t an y one factor i s not compared to all other factors simultaneously but rather one at a time. For an easy explanation let us take n = 3 in the following example .

QUANTITATIVE TECHNIQUE FOR DECISIO N MAKING IN CONSTRUCTION

Three factors: 1 , 2 and 3. Factor 1 is twice as important a s Factor 2. Factor 2 is three times as important as Factor 3. Factor 1 is six times as important as Factor 3. Then, these scores can be entered into a matrix A as follows:

A = 1 2 3

1 1

1/2 1/6

2 2 1

1/3

3 6 3 1

Matrix A is called th e "reciproca l matrix " because al l its lower triangula r elements ar e equal t o the reciprocal o f its upper triangula r elements . Th e maximum eigenvalue X of matrix A can be found by solving

I A- Xl I = 0

The eigenvecto r X , corresponding t o th e maximu m eigenvalu e A, , which satisfies tha t AX = AX, is found t o be:

X = 0.6 0.3 0.1

- Factor 1 - Factor 2 - Factor 3

The higher th e value of the element in X, the more important th e factor is .

1.2.2 Reader s may find i t difficul t t o understand wh y th e resultant eigenvecto r can represent importance. The following gives a brief explanation. Details of it can be found i n the work of Saaty (1977 and 1980) .

Suppose a set of n objects are to be compared in pairs and thei r individua l importance (assumed known) are denoted by xv x 2, , xn, then the pairwise comparison ma y be represented by a (n x n) square reciprocal matrix A as follows:

1 2 n 1 2

A =

1 X 2 / X l

x/x2

1 x/x

1 n

x/x 2 n

x /x. x /x. n 1 n 2

ANALYTIC HIERARCHY PROCESS I

If a column vecto r X is defined suc h tha t x v x 2, .... , xn (i.e . the individua l importance o f objects 1 , 2, ..., n) are the elements of X such tha t

X =

then,

AX =

* 2 / X l

V*2

x Ix, x lx^ . n 1 n 2

x/x " 1 n

x/x I n

l

r x' i X2

L X n J

xx +x x + +x x

x2 + x2 + + x2

X + x + + x n n n

= n

= nX

Therefore, AX = nX, and n is, by definition, a n eigenvalue of matrix A. The matrix X is the solutio n eigenvecto r o f AX = XX for takin g ^=n . n is , by Perron-Frobenius Theorem , th e maximum eigenvalu e o f A. This explain s why the eigenvector contains the ranking of the importance of the n objects.

1.2.3 Th e above theory is based on the assumption tha t all entries of the matrix A are consistent which means that (a..)(a. k) = a.k. The example in Section 1.2. 1 is a consistent case . Such perfec t consistenc y i s possible onl y i f we ca n

QUANTITATIVE TECHNIQUE FOR DECISIO N MAKIN G IN CONSTRUCTION

construct matrix A based on the weightings of individual objects (i.e . xv x 2, ..., xn). However, in the application of AHP, this will not be the case, that is, one can only construct matrix A first by pairwise comparisons and then fin d out the values of xv x 2,...., xn. This will create inconsistency in the reciprocal matrix A. Just tak e a new example of three footbal l teams : if team 1 beats team 2 by 2:1 and team 2 beats team 3 by 3:1, it is not necessarily the case that team 1 will beat team 3 by 6:1 (this is exactly the case in the example in Section 1.2.1) , and if not, inconsistency occurs .

When inconsistency occurs, the problem AX = nX becomes AX = A,maxX. For the reciprocal matri x A, ^max will not be equal t o n. I t has been proved b y Saaty tha t ^ max is closer t o n when matri x A is closer t o consistency . A is consistent i f an d onl y i f X = n. X i s alway s greate r tha n n i f A i s

J ma x ma x J o

inconsistent. The further X i s from n, the more the inconsistent the matrix max

is. Let us now modify th e example in Section 1.2. 1 t o an inconsistent cas e as follows: Three factors: 1 , 2 and 3. Factor 1 is twice as important as Factor 2. Factor 2 is three times as important as Factor 3. Factor 1 is four time s as important as Factor 3.

The reciprocal matrix is therefore written as:

A = 1

1/2 1/4

2 1

1/3

4 3 1

The maximum eigenvalue is 3.018 and the corresponding eigenvector X (or called priority vector) is :

x =

0.56 0.32 0.12

Readers can now see the difference o f this priority vector and th e priorit y vector in Section 1.2.1 .

It is important t o note tha t the summation o f all the elements in a priority vector i s equa l t o 1 (se e th e tw o previou s examples) . Thi s i s calle d

ANALYTIC HIERARCH Y PROCESS I

"normalization" o f th e priority vector . Th e normalization i s necessary i n order t o ensure uniqueness o f the vector. The two priority vectors we have seen are said to be normalized additively , tha t is, the elements add up to 1. There are, however, cases that need a priority vector to be normalize d multiplicatively. W e will see such example s in Section 2. 3 in Chapte r 2 . The evaluation o f the maximum eigenvalu e of a n x n matrix and hence its corresponding eigenvecto r (i.e . the priority vector) ca n be easily found i n many mathematics books and software o n the market .

Table 1. 1 shows the scales of 1 to 9 as recommended by Saaty for inputtin g values into the reciprocal matrix .

Intensity of relative importance

1

3

5

7

9

2,4,6,8

Definitions

Equal importanc e

Moderate importanc e of one over anothe r

Essential or stron g importance

Demonstrated importance

Extreme importanc e

Intermediate value s between the tw o adjacent judgement s

Explanation

Two activities contribut e equally t o the objective s Experience and judgemen t slightly favoured one activit y over anothe r Experience and judgemen t strongly favoured on e activity ove r anothe r An activit y i s strongly favoured an d it s dominanc e is demonstrated i n practic e The evidence favouring on e activity ove r anothe r i s of the highest possibl e order o f affirmation When compromis e i s needed

Table 1.1 Scale s of 1 to 9 for pairwis e comparisons (Saat y 1977 )

Three Levels of Hierarch y

So far, only how n object s ar e ranked based on a single objective ha s been discussed. Now , a hierarchy o f 3 levels (Fig . 1.1 ) i s looked into . The firs t

QUANTITATIVE TECHNIQUE FOR DECISION MAKING IN CONSTRUCTION

level ha s a single goa l (o r principa l objective) . Th e secon d leve l ha s m subordinate objective s (o r factors ) an d thei r ranking s ar e derive d fro m pairwise comparisons based on the goal of the first level . The third level has n objects (o r tenderers) which are to be ranked. The problem is to determine how well the objects meet the goal through the intermediate second level of subordinate objectives . The procedures are described below.

Level 1

Level 2

Level 3

Principal Objective (1) (Best Tender)

Subordinate Objectives (m) (Factors)

Objects (n) (Tenderers)

Fig. 1. 1 A 3-leve l hierarch y structur e

Step 1 Construc t a reciprocal matrix B of dimension m x m with pairwise comparisons of the factors with respect to the principal objectiv e as the elements of B. Then find the priority vector of matrix B and denote i t by Xb.

Step 2 Sinc e there are n tenderer s t o be ranked, a total of m reciproca l matrices A. (i = 1, 2, ..., m) of size n x n are to be formed, eac h of which consist s o f element s o f pairwis e comparison s o f th e tenderers with respect to a single factor a s objective, tha t is:

A = 2

n

Using factor 1 as objective


A2 = Using factor 2 as objective

A =

n

Using factor m as objective

Step 3 I f X ( i = 1, 2,..., m) is the priority vector of the corresponding A., then an n x m composite matrix C can be formed b y taking X as columns in C in sequence such that :

C = [Xp x2,..., XJ

Step 4 Th e resultant priority vector, Xc, is the result of the multiplication of matrices C and Xb, that is:

X = C x Xu c b

From the result of Xc, the ranking of the tenderers can be obtained. These 4 steps will be further illustrate d in the next section .

4 Evaluatio n of Tenders GOAL

F1 F 2 F 3 F 4 F 5 F 6 F 7 F 8 F 9 F1 0 F1 1 F1 2 F1 3

i.e., m=13 & n=4

Fig. 1.2 3-leve l hierarch y fo r evaluation o f tender s

i 8 QUANTITATIV E TECHNIQUE FOR DECISION MAKING IN CONSTRUCTION

A tender evaluation exampl e is shown as a 3-level hierarchy AHP problem (Fig. 1.2) . Fou r tenders , A, B, C and D, are to be evaluated. They are fro m four differen t consultin g firms who wish to bid to undertake th e planning, design and supervision o f construction fo r a construction project . So , four tender proposals have been submitted to the client. There are thirteen factor s (under five headings) t o be considered by the client. These factors are:

Consultant's Experienc e F1: Relevan t experience and knowledge . Response to The Brief F2: Understandin g o f objectives . F3: Identificatio n o f key issues. F4: Appreciatio n o f project constraint s and special requirements. F5: Presentatio n o f innovative ideas. Approach to Cost-Effectivenes s F6: Example s and discussion of past projects to demonstrate the consultant's

will and ability to produce cost-effective solutions . F7: Approac h t o achieve cost-effectiveness o n this project . Methodology an d Work Programme F8: Technica l approach . F9: Wor k programme and project implementation programme . F10: Arrangements for contract management and site supervision . Staffing F l l : Organizatio n structure . F12: Relevant experience and qualification o f key staff. F13: Responsibilities and degree of involvement of key staff.

The matrix B (see step 1 of Section 1.3) i s shown in Fig. 1.3. Its elements are pairwise comparisons o f the factors with respect to the principal objective . For example, Fl an d F2 are of equal importance, and therefore elemen t 1- 2 of matrix B is entered a s 1 . F3 is of moderate importanc e ove r F10 , an d therefore elemen t 3-10 is entered as 3.


Goal F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13

F1 F 2 F 3 F 4 F 5 F 6 F 7 1 1 1 1 1 1 1 / 5

1 1 1 1 1 1 / 5 1 1 1 1 1 / 5

1 1 1 1 / 5 1 1 1/ 5

1 1/ 5 1

All lowe r triangula r element s =a.

ji =1/a.

ij

F8 1/7 1/7 1/7 1/7 1/7 1/7 1/5 1

F9 1/3 1/3 1/3 1/3 1/3 1/3 4 6 1

F10 3 3 3 3 3 3 7 9 5 1

F11

5 7 3

1/3 1

F12 F1 3 1/7 1/ 7 1/7 1/ 7 1/7 1/ 7 1/7 1/ 7 1/7 1/ 7 1/7 1/ 7 1/5 1/ 5 1 1

1/6 1/ 6 1/9 1/ 9 1/7 1/ 7 1 1

1

Fig. 1.3 Matri x B for tender evaluation

The priority vector Xb of matrix B is:

x> =

0.0258 0.0258 0.0258 0.0258 0.0258 0.0258 0.1074 0.2136 0.0586 0.0126 0.0258 0.2136 0.2136

The next step is to construct 1 3 matrices of size 4 x 4, each of which consists of elements of pairwise comparisons of the 4 tenders with respect to a single factor as objective. It is impossible to show all the 13 matrices here and only A, A2 and A13 are shown below:

QUANTITATIVE TECHNIQUE FOR DECISIO N MAKING I N CONSTRUCTION

A = A B C D

A 1 1 1 1/2

B 1 1 1 1/2

C 1 1 1 1/2

D 2 2 2 1

Using Factor 1 as objective

A = A B C D

A 1 2 1/2 1

B 1/2 1 1/4 1/2

C 2 4 1 2

D 1 2 1/2 1

A

»- B C D

A 1 1/2 1/5 1/5

B 2 1 1/3 1/3

C 5 3 1 1

D 5 3 1 1



The priority vectors, Xp X2,...., X13, of the matrices, Av A 2,...., A13 respectively are:

* , -

0.2857 0.2857 0.2857 0.1428

* 2 =

0.2222 0.4444 0.1111 0.2222

* B -

0.5183 0.2839 0.0989 0.0989

Hence, a composite matrix C can be formed a s follows. Not e tha t the 1st , 2nd and 13t h columns of C are Xp X 2 and X 3 respectively.

C =

0.2857 0.222 2 0.518 3 10.2857 0.444 4 0.283 9 0.2857 0.111 1 0.098 9 |o.l428 0.222 2 0.098 9

ANALYTIC HIERARCH Y PROCES S I

The ranking of the four tender s = C x Xb

0.3548" 0.2599 0.0851 0.3002

Therefore, th e best tender is A, which scores 0.3548; the second best tende r is D, which scores 0.3002; and so on.

Four Levels Hierarchy — Tender Evaluatio n

A 3-level hierarchy has just been discussed . Th e following wil l illustrate a more complicated tender evaluation example in which a 4-level hierarchy is involved (Tang , 1995) .

There are four international contractors tendering for a large civil engineering design-and-build contrac t involvin g earth work, road work, tunne l work , building wor k an d E& M work . Eac h o f thes e works form s a part o f th e contract but each is significant enoug h to form a contract by itself if it is not a larg e turn-ke y internationa l contrac t requirin g hig h standar d o f workmanship and reliability. Hence, the selection of tender must be carried out with exceptional care. Fig. 1.4 shows a 4-level hierarchy, following which the client evaluates the tenders.

Management

ability

Fig. 1.4 4-leve l hierarch y fo r evaluation o f tender s


The procedures of evaluating a 4-level hierarchy are described as follows:

Step 1 Construc t a reciprocal matrix P of dimension 5 x5 wit h pairwise comparisons o f the different work s with respect t o the principa l objective a s the element s o f P. The higher th e percentage o f the work in the contract, the more important tha t work is. Then fin d the priority vector of matrix P and denote it by Pb.

Step 2 Construc t five reciprocal matrices Q. (i = 1, 2, 3, 4, 5) of size 6 X 6, each of which consists of elements of pairwise comparisons of the six factors with respect to a particular work as objective.

Step 3 I f X ( i = 1, 2, 3, 4, 5) is the priority vector of the correspondin g Q., then a 6 x 5 composite matri x R is formed b y takin g X a s columns in R in sequence such that :

R = [Xv X 2, X3, X4, X5]

Step 4 Construc t si x reciprocal matrices S . (i = 1,2, 3 , 4, 5 , 6) o f size 4 x 4 , eac h of which consist of elements of pairwise comparison s of the four tenderer s with respect to a single factor a s objective.

Step 5 I f Y. (i = 1, 2, 3, 4, 5, 6) is the priority vector of the corresponding S., then a 4 x 6 composite matrix T can be formed by taking S. as columns in T in sequence such that :

T = [ Y Y Y Y Y Y 1 1' 2 ' 3 ' 4 ' 5 ' 6

Step 6 Th e resultant priority vector, Xc, is the result of the multiplication of matrices T, R and Pb, that is:

X = T x R x Ph c b

Similar to the previous example, the ranking of the tenderers can be known after th e result of X i s obtained.

c


More Examples on Applications o f AHP

Example 1. 1 A3 - level hierarchy decision problem

Level 1 Goal: Select the best wastewater treatmen t alternativ e

Level 2 Factors for consideration : 1. Sewag e flow 2. Influent/Effluen t standar d 3. Siz e of site 4. Natur e of site 5. Lan d cost 6. Loca l money for constructio n 7. Foreig n money component fo r constructio n 8. Loca l skill for constructio n 9. Communit y suppor t 10. Powe r source 11. Availabilit y of local material 12. Cos t of operation and maintenanc e 13. Professiona l skil l available for operation and maintenanc e 14. Loca l technical skill available for operation and maintenanc e 15. Administratio n set-u p 16. Trainin g 17. Professiona l ethic s 18. Climat e 19. Loca l water-borne disease s 20. Endemi c vector-borne (water-related ) disease s


Level 3 Wastewater treatmen t alternatives : 1. Stabilizatio n pond s 2. Full y aerated lagoons + Secondary settlemen t 3. Fully/Partiall y aerated lagoon s 4. Simpl e percolating filtratio n 5. Modifie d percolatin g filtratio n 6. Conventiona l activated sludge process 7. Deep-shaft/High-purit y oxyge n processes 8. Primar y settlemen t 9. Lan d applicatio n 10. Rotatin g biological contactor s 11. Oxidatio n ditche s 12. Packag e activated sludge plants 13. Packag e high-purity oxygen plants

Example 1. 2 A 4-level hierarchy decision proble m

Level 1 Goal: Select the optimal alignment fo r a road projec t

Level 2 Design aspects: 1. Projec t su m 2. Projec t duratio n 3. Constructio n risk s 4. Operatio n and maintenanc e 5. Environmenta l impac t

ANALYTIC HIERARCH Y PROCES S I

Level 3 Construction methods : 1. Tunne l 2. Bridg e 3. Immerse d tub e

Level 4

Alignment alternatives : 1. Alignmen t 1 2. Alignmen t 2 3. Alignmen t 3 4. Alignmen t 4

Example 1. 3 Anothe r 4-level hierarchy decision proble m

Level 1

Level 2

Goal

Level 3 i

Level 4

Level 1

Goal: Selec t the most appropriate contractual arrangement for a construction project

Level 2

Components of the project :

1. Earthwor k 2. Roadwor k 3. Drainag e and sewerage work 4. Buildin g work 5. Utilit y work

i 1 6 QUANTITATIV E TECHNIQUE FOR DECISION MAKING IN CONSTRUCTION

Level 3

Factors affecting th e choice of contractual agreement : 1. Pro j ect definitio n 2. Owne r preference s 3. Publi c laws 4. Curren t market condition s 5. Projec t locatio n 6. Projec t financin g 7. Schedul e 8. Assumptio n o f risks 9. Scop e of work 10. Duratio n o f work

Level 4

Contractual arrangement alternative s 1. Cos t plus contrac t 2. Targe t price contrac t 3. Uni t price contrac t 4. Lum p sum contrac t

1.7 Advantage s o f Usin g AHP i n Decisio n Makin g

The advantages of using AHP are as follows:

1.7.1 I t provides a systematic procedure for comparisons between objects under a large number of factors. It facilitates the employment of subjective weighing of objects based on experience .

1.7.2 Beside s quantifiabl e factor s (e.g . tender sum , projec t duration , etc.) , th e method enables the consideration of unquantifiable/subjective factor s which are important in decision making processes.

1.7.3 Th e size and th e complexity o f a problem ca n be broken dow n into smal l items (or called clusters) for analysis. For the number of levels of the hierarchy is flexible dependin g on the size and the requirements of the problem.

1.7.4 Th e resultant priority vector obtained from this method can give an indication of how much one object is better than another. This can hardly be achieved by intuition alone .


Exercise Question s

Question 1

Explain th e consistenc y an d inconsistenc y o f a reciprocal matri x i n th e Analytic Hierarchy Proces s (AHP) . How is the matrix' s larges t eigenvalu e (A,max) related to the consistency?

In th e applicatio n o f AHP, what typ e o f reciprocal matri x (consisten t o r inconsistent) i s usually used? Why?

Question 2

Construct a 5-level hierarchy problem related t o construction, th e 1s t level being the goal and the 5th level being the alternatives for selection .

BIBLIOGRAPHY

Chapters 1 and 2

Barzilai, J. (1997) . "Deriving weights from pairwise comparison matrices." Journal of the Operational Research Society, 48, 1226-1232.

Barzilai, J. and Golany, B. (1994). "AHP rank reversal, normalization and aggregation rules." INFOR, 32 (2) , 57-64.

Belton V and Gear , T. (1983) . "O n a short-coming o f Saaty's method o f analyti c hierarchies." Omega, 11 (3), 228-230.

Donegan, H.A., Dodd, EJ. and McMaster, T.B.M. (1992). "A new approach t o AHP decision-making." The Statistician, 41, 295-302.

Donegan, H.A. Dodd, EJ. and McMaster, T.B.M. (1995). "Theory and methodolog y — inverse inconsistency i n analytic hierarchies." European Journal of Operational Research, 80, 86-93.

Johnson, C.R . Beine , W.B. and Wang,T.J . (1979) . "Right-lef t asymmetr y i n a n eigenvector ranking procedure." Journal of Mathematical Psychology, 19, 61-64.

Saaty, T.L. (1977). "A scaling method for priorities in hierarchical structures." Journal of Mathematical Psychology, 15 , 234-281.


Saaty, T.L. (1980). The Analytic Hierarchy Process . McGraw-Hill, New York.

Tang, S.L . (1995) . "Tende r evaluatio n usin g Analytic Hierarch y Process. " Pape r presented in the 1st Intexnational Symposium on Project Management, Northwestern Polytechnical University, Xi'an, China, September 1995 .

Tung, S.L. (1997). Right and left eigenvector inconsistency in analytic hierarchy process. M.Sc. Dissertation, Civi l and Structura l Engineering Department , The Hong Kong Polytechnic University .

Tung, S.L. and Tang, S.L. (1998) . "A comparison o f the Saaty's AHP and Modifie d AHP for righ t and left eigenvecto r inconsistency." European Journal of Operational Reseaxch, 106 , 123-128.

Vargas, L.G. (1982). "Reciprocal matrices with random coefficients." Mathematical Modelling, 3, 69-81.

Chapters 3 , 4 , 5 , 6 and 7

Levin, R.I. , Rubin , D.S. , Stinson, J.P. and Garde n Jr., E.S . (1992) . Quantitative Approaches to Management. 8t h Edition. McGraw-Hill , New York, USA.

Lind, DA. and Mason, R.D. (1997). Basic Statistics: for business and economics. 2nd Edition. Irwin , Times Mirror Higher Education Group , USA.

Meredith, D.D., Wong, K.W., Woodhead, R.W. and Wortman, R.H. (1985). Design and Planning of Engineering Systems. 2n d Edition. Prentice-Hall , New Jersey, USA.

Pilcher, R. (1992) . Principles of Construction Management. 3rd Edition . McGraw -Hill, London, U.K.

Tang, S.L. and Poon, S.W. (1987). Project Management, Vol. 2. Education Technology Unit, Hong Kong Polytechnic, Hong Kong.

Chapter 8

Bather, J. (2000) . Decision Theory: An Introduction to Dynamic Programming and Sequential Decisions. John Wiley & Sons, New York, USA.

Bellman, R.E. (1957). Dynamic Programming. Princeton University Press, New Jersey, USA.

Jackson, W. (1995) . Optimization. University of London, London, U.K.

BIBLIOGRAPHY

Makower M.S. and Williamson, E. (1985). Operational Research. 4th Edition. Hodder and Stoughton Education, UK.

Ozan, T.M. (1986). Applied Mathematical Programming for Engineexing and Production Management. Prentice-Hall , New Jersey, USA.

Chapters 9 and 1 0

AbouRizk, S . M. and Halpin , D . W. (1991) . "Visua l Interactiv e Fittin g o f Bet a Distributions" Journal of Construction Engineexing and Management, ASCE, 117 (4), 589-605.

AbouRizk, S . M. and Halpin , D . W. (1994) . "Fittin g beta distribution s base d o n sample data." Journal of Construction Engineering and Management, ASCE, 120 (2), 288-305.

AbouRizk, S.M. (2000). Simphony CYCLONE user's guide, Construction Engineerin g & Managemen t Program , University of Alberta, Canada .

Fente, J. , Schexnayder , C . an d Knutson , K . (2000 ) "Definin g a probabilit y distribution function fo r construction simulation." Journal of Construction Engineering and Management, ASCE, 126 (3), 234-241.

Lichtenstein, S., Fischhoff, B. and Phillips, L.D. (1977). "Calibration of probabilities: the stat e o f th e ar t t o 1980. " Judgment under uncertainty: heuristics and biases. Cambridge University Press, U.K. 306-334.

Halpin, D.W. (1977). "CYCLONE-A method for modeling job site processes." Journal of Construction Division, ASCE, 103 (3), 489-499.

Halpin, D.W. and Riggs, L. (1992). Planning and analysis of construction operations. Wiley, New York, USA.

Law A. and Kelton D. (1982). Simulation Modeling and Analysis. McGraw Hill, New York, USA.

Lu, M. (2002). "Enhancing PERT simulation through ANN-based input modeling." Journal of Construction Engineering and Management, ASCE, 128 (5), 438-445.

Lu, M., and Anson, M. (2004). "Establish concrete placing rates using quality control records from Hon g Kong building construction projects. " Journal of Construction Engineering and Management, ASCE. Vol. 130, March/Apri l Issue.

Pidd, M. (1989). Computer Simulation in Management Sciences. 2n d ed. Wiley, New York, USA.


Chapters 1 1 and 1 2

Kroenke, D., and Hatch, R. (1994). Management of Information Systems, 3rd edition , McGraw-Hill, Singapore .

Tenah, K. A. (1984). "Management o f Information i n Organizations and Routing" . Journal of Construction Engineexing and Management, ASCE, 110(1), pp. 101-118 .

Antill, J.M. an d Woodhead , R.W . (1990) . Critical Path Methods in Construction Practice. 4th ed. , Wiley, NY.

Tang, S.L., Poon, S.W, Ahmed, S.M. and Wong, K.W. (2003). Modern Construction Project Management, 2nd ed. , Hong Kong University Press, Hong Kong.

ABOUT THE

AUTHORS

S.L Tang S.L. Tang is a faculty membe r i n th e Departmen t o f Civi l and Structura l Engineering o f the Hong Kon g Polytechnic University . H e is a Chartere d Civil Engineer and obtained his B.Sc. in Civil Engineering from the University of Hong Kong in 1972, M.Sc. in Construction Engineering from the National University o f Singapor e i n 1977 , and Ph.D . fro m th e Civi l Engineerin g Department o f Loughborough University , U.K. in 1989 . Dr Tang had abou t seven years of working experience in civil engineering practice in contracting/ consulting firm s an d government department s befor e h e joined th e Hon g Kong Polytechnic University . He is currently involved in the teaching an d research o f constructio n management , an d wate r an d environmenta l management, an d has written ove r ninety journal/conference paper s an d books related to the area of his expertise. As a visiting scholar, Dr Tang taught in the civil engineering and construction management departments of several universities, such as University of Technology Sidney of Australia, Tsinghua University, Tongji University, Zhejiang University , and Huaqiao Universit y of China, and Florida International University of USA.

Irtishad U. Ahmad Irtishad U . Ahmad, RE . is th e Chai r o f th e Departmen t o f Constructio n Management at the Florida International University (FIU) since January 2004. He was th e Graduat e Progra m Directo r i n th e Departmen t o f Civi l an d Environmental Engineerin g a t FIU from 199 8 to 2003 . He is a registered professional enginee r in the state of Florida. Dr. Ahmad earned his Ph.D. in Civil Engineering from th e University of Cincinnati, Ohio, USA in 1988 . He taught previously in North Dakota State University, University of Cincinnati, and Bangladesh University of Engineering and Technology . Dr Ahmad is a consultant and a practising engineer in the areas of structural analysi s and design, reinforced concret e design, and forensic engineering . He has served


in severa l valu e engineerin g team s a s a cos t engineerin g an d projec t management expert . He is a frequent speake r in national and internationa l conferences o n topics ranging from information technolog y in constructio n to effects o f wind loading on structures. Dr Ahmad completed several funde d research projects related to construction and he published extensively. He is the curren t Editor-in-Chie f o f the Journal o f Management i n Engineerin g published by the American Society of Civil Engineers.

Syed M. Ahmed Syed M. Ahmed is a faculty membe r and Graduate Program Director in the Department o f Constructio n Managemen t a t th e Florid a Internationa l University i n Miami , Florida US A since December , 1999 . Prior t o thi s h e was a n academi c staf f membe r i n th e Departmen t o f Civi l & Structura l Engineering o f the Hong Kong Polytechnic University fo r ove r four years . He has a B.Sc. (Hons. ) degre e in Civi l Engineering fro m th e University of Engineering and Technology, Lahore in 1984, and M.S. and Ph.D. degrees in Civil Engineering (majorin g i n Construction Managemen t with a minor in Industrial and Systems Engineering) from the Georgia Institute of Technology, Atlanta, Georgia , USA in 198 9 and 199 3 respectively. Dr Ahmed's research interest is in the areas of total quality management, construction safety an d quality, contract management an d risk analysis , construction procuremen t and financia l management , informatio n technology , and construction an d engineering education. H e has published extensively in refereed internationa l journals an d conference s an d has over ten years of construction industr y experience.

Ming Lu Ming L u i s a faculty membe r i n th e Departmen t o f Civi l and Structura l Engineering o f th e Hon g Kon g Polytechni c University . H e has workin g experience on the planning and construction of numerous civil and industrial projects i n th e past . I n 2000 , he earned hi s Ph.D. degree in Constructio n Engineering & Managemen t fro m th e University o f Alberta, Canada . Hi s current research interest is in the area of computational construction projec t management aide d by artificial intelligenc e an d compute r simulations . I n recent years, Dr. Lu has developed the novel approaches for neural networks modelling, resource schedulin g an d operation s simulatio n fo r enhancin g efficiency an d productivit y o f comple x constructio n systems . Thes e developments hav e been published i n the ASCE journals o n constructio n engineering, management, and computing methods.

quantitative techniques for decision making in … · quantitative technique for decision making in...

Documents