Time-cost Optimization And Its Ant Colony Algorithm In Network Plan

Time-cost optimization is considered as a highly important content to the scheduling of the project management. A suitable scheduling arrangement can achieve obvious benefit for the project during the construction. The research about time-cost optimization problems in the early stage is mostly based on mathematical programming methods and heuristics methods,but such kinds of methods have many drawbacks. In order to overcome these limitations,ant colony algorithm technique is odopted in the schedule optimization due to its incomparable properties in combinatorial optimization,and so a more precise and effective multicriteria optimal approach can be obtained.For time-cost optimization problems in which it is continuous relationship between time and cost of activaty, most of the current research projects focus on solving the minimum cost or minimum cost period. An improved method is proposed based on the idea of ant colony algorithm for continuous space optimization using grading method,in which the lowest cost is considerd as the objective function. For time-cost optimization problems with time limit,in this paper,penalty function is adopted to deal with time constraint condition,which is effective to solve the problems hard to deal with in algorithm,and unifies the solving method of two kinds of problem(with and without time constraint).For time-cost optimization problems in which it is discrete relationship between time and cost of activaty, most of the current research projects focus on solving time-cost trade-off curve. A corresponding method is proposed in which time objective and cost objective are required to be minimum. In the method,the adaptive weight approach is adopted to integrate the two objectives of time and cost into a single objective,and ant colony algothem is applied to search for pareto solutions.The two types of models are all illustrated with case studies. The final results demonstrate not only the accuracy and highly effective capability of the models,but also their practical application value.