Keywords: discrete event simulation, genetic algorithms, time-cost trade-off, multiobjective optimization, construction operations modelling, computer simulation.

A construction project is composed of a network of activities. To perform the activities of a project, planners usually have to decide the construction technologies, operation processes and the associated resources, that are to be used in the operations, including crew sizes and equipment from different alternatives. However, different construction methods and the related resource combinations for performing an activity create various durations and costs for that particular activity and therefore influence the total cost and duration for a project. Hence, planners have to face the decision of finding the most cost effective way to complete a project within the desirable duration for a project. These decisions are usually made based on the so-called time-cost trade-off (TCT) analysis. The purpose of such analysis is to reduce the project cost but not to impact on the project duration. The results of TCT analysis provide a TCT curve that the associated duration and cost for each resource combination can be observed and ultimately the balance of time and cost can be optimized.

Intensive research has been dedicated in solving TCT problems for decades. The models developed in the past can be classified according to the limited conditions when they were used in solving TCT problems. The first type of TCT model is limited in that it only can be used when a linear relationship exists between time and cost of an activity and in addition, the curve showing the time-cost relation of an activity are composed by discrete pairs. Most models were established based on these assumptions. For example, Perera [1] and Demeulemeester et al. [2] used linear programming and dynamic programming to solve TCT problems. Liu et al. [3] presented the hybrid mechanism for integrating linear programming and integer programming to tackle the TCT problem. Mathematical programming techniques can exactly reach optimal solution for TCT problems but require a great computation effort. Therefore, different models such as [4] were proposed using GAs to analyze TCT problems. The second type of TCT model assumes that the relationship between the time and cost of construction activities can be represented by a continuous curve. The nonlinear nature of time-cost trade-off relations was recognized by Fulkerson [5] and Berman [6] during the 1960s. However, until recent decades, the study of such types of model were still to be explored. For instance, Li et al. [7] presented a computer system integrating the machine learning and the genetic algorithms to tackle this type of TCT problem.

Many construction alternatives, especially those for the new construction methods, have not been applied in the project before they are selected as construction method. It is risky to directly estimate their time and cost for a project. However, such construction methods and associated resource combinations can be modeled by discrete event simulation sophisticatedly and their impact on time and cost of a project verified. Judging from literature reviews, there has been little reported research focusing on solving the TCT problem using discrete event simulation techniques. On the other hand, however, simulation techniques have to perform an exhaustive search to verify the impact of system performance for all possible resource combinations. This paper presents a mechanism that integrates genetic algorithms (GAs) and discrete event simulation technique to solve TCT problems. The construction operations are modeled using discrete event simulation techniques to obtain the duration and cost of a project. Then, the impact on time and cost for possible resource combinations are verified and screened by the GA. A case study shows that the TCT curve can be efficiently located by a GA-simulation mechanism. Additionally, a computer program is implemented to automate the execution of this new approach, which provides a practical tool for a construction project planner to apply the GA-simulation mechanism to practice.

1

S. Perera, "Linear programming solution to network compression", J. of Constr. Division, ASCE, 106, 315-327, 1980.

2

E. L. Demeulemesster, W. S. Herroelen, S. E. Elmaghraby, "Optimal procedures for the discrete time/cost trade-off problem in project networks", European J. of Operational Research, 88, 50-68, 1996. doi:10.1016/0377-2217(94)00181-2

3

L. Liu, S. A. Burns, and C.-W. Feng, "Construction time-cost trade-off analysis using LP/IP hybrid method", J. of Constr. Engrg. And Mgmt., ASCE, 121(4), doi:10.1061/(ASCE)0733-9364(1995)121:4(446)

4

X. M. Zheng, S. T. Ng, and M. M. Kumaraswamy, "Applying a genetic algorithm-based multiobjective approach for time-cost optimization", J. of Constr. Engrg. and Mgmt., ASCE, 130(2), 168-176, 2004. doi:10.1061/(ASCE)0733-9364(2004)130:2(168)

5

D. Fulkerson, "A network flow computation for project cost curves", Mgmt. Science, 7, 167-178, 1961. doi:10.1287/mnsc.7.2.167

6

F. B. Berman, "Resource allocation in a pert network under continuous activity time-cost functions", Mgmt. Science, 10, 724-735, 1964. doi:10.1287/mnsc.10.4.734

7

H. Li, J.-H. Chao, and P. E. D. Love, "Using machine learning and GA to solve time-cost trade-off problems", J. of Constr. Engrg. and Mgmt., ASCE, 125(5), 347-353, 1999. doi:10.1061/(ASCE)0733-9364(1999)125:5(347)

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £105 +P&P)