Keywords: overall benefit-duration optimization, MATLAB, genetic algorithm,.

Given a normally-scheduled project network with a set of activities to be completed according to their precedence relationships, the schedule can be compressed such that opportunity income exceeds the cost increment incurred by network compression. The objective is to make a sequence of decisions to crash activities on the network and hence, compress the schedule to a desired limit where the optimal overall economic benefit of owner is reached. This problem is referred to as overall benefit-duration optimization (OBDO) [1]. Unlike previous schedule optimization models that have focused on project durations or costs minimization from contractors' perspective, the authors present a MATLAB-programmed genetic algorithm (GA) solution for the sake of maximizing owner's economic benefit [2].

In this paper, objective function of OBDO model is formulated. A test example and its GA application in MATLAB are illustrated to prove the feasibility and practicability of OBDO concept. In the case study, GA is employed to improve scheduling optimization efficiency through network compression. As a computer programming language and a software environment for using that language effectively, the interactive environment of MATLAB allows GA to manage OBDO objective function, variables, import and export data, perform calculations, generate plots and figures, and develop and manage files for use. The MATLAB- programmed GA solution goes on by performing reproduction, crossover and mutation to initial string population representing the compressed schedule of the project network. Through crossover and mutation, the strings are evaluated by fitness function (objective function of OBDO) for the sake of selecting better solutions for the next generation. After several generations, the best member of the population turns out to represent an optimal solution until the owner's overall benefit is maximized. For the final output, the compressed schedule and owner's optimal benefit are gathered and exported for final decision-making. With the motivation of learning from efficiency and the capability of solving complex optimization problems, the MATLAB-programmed GA solution overcomes the limitations of traditional methods, provides a synergy with more problem-solving power.

1

H. Pan, S.K. Ting, "A Genetic Algorithms Solution to Overall Benefit-duration Trade-off for Project Owners", Proceeding of 4th Asia-Pacific Simulated Evolution and Learning Conference, SEAL'02, Singapore, 2002, 5pp.

2

S.K. Ting, H. Pan, "Overall Benefit-Duration Optimization (OBDO) for Owners in Large-Scale Construction Projects", submitted to ASCE for publication, 2003.

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 £82 +P&P)