A Hybrid Artificial Intelligence Approach for Optimizing Construction Time-Cost Tradeoff

• Main Authors: Phu-Cuong Cao, 高富強
• Other Authors: Min-Yuan Cheng
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/58113155785211078591

碩士 === 國立臺灣科技大學 === 營建工程系 === 98 === Problems in construction industry are complex, full of uncertainty, and vary based on site environment. In the management of a construction project, the project duration can often be compressed by accelerating some of its activities at an additional expense. This is the so-called time-cost tradeoff (TCT) problem, which has been studied extensively in the project management literature. TCT decisions, however, are complex and require planners to select appropriate resources for each project task, including crew size, equipment, methods, and technology. As combinatorial optimization problems, finding optimal decisions is difficult and time consuming considering the number of possible permutations involved. The present study applies a new optimization approach, named K-means clustering with Chaos Genetic Algorithms (KCGA) proposed by Cheng and Huang (2009), to solve the TCT problem, that is to minimize the total project cost as an objective function and account for project-specific constraints on time and costs. To improve existing methods, particularly to demonstrate how the genetic algorithms (GA) optimizer can be improved by incorporating a hybridization strategy, KCGA employs the chaos procedure to maintain the population diversity of GA and K-means clustering technique to speed up the optimization search in GA. Besides, the TCT-KCGA model is capable of treating all existing types of activity time-cost functions, such as linear, nonlinear, discrete, discontinuous, and a hybrid of the above; and being insensitive to the scales of time and cost. Through two experimental studies, it can be revealed that the hybrid KCGA approach is able to reduce the computational amount and improve estimation accuracy when compared to other algorithms separately. On the whole, the KCGA model is shown effective and efficient in conducting advanced time-cost analysis. Future applications of the proposed TCT model are suggested in the conclusion.