基于改进遗传算法的关键链项目进度计划优化  

Critical Chain Project Schedule Optimization Based on Improved Genetic Algorithm

作　　者：张琦[1] 廖良才[1] 王卫威[1] 

机构地区：[1]国防科学技术大学信息系统与管理学院,湖南长沙410073

出　　处：《计算机技术与发展》2014年第4期1-5,11,共6页Computer Technology and Development

基　　金：国家自然科学基金资助项目(61201337)

摘　　要：项目进度计划一直是项目管理中的重点和难点,其原因是在资源和工序约束的双重约束下,项目进度计划的求解是一个NP难问题,尤其是在多资源约束下更增加了问题的复杂度。文中从关键链这一角度出发,分析了该问题并建立模型。应用自适应遗传算法对模型进行求解,对遗传算法的编码、选择、交叉、变异等操作进行了设计,并在此基础上对传统的遗传算法进行改进。最后,通过案例的求解,验证了关键链方法在研究多资源约束项目进度计划问题上的可行性,以及自适应遗传算法在求解该问题上的优越性。Project schedule has been a focus and difficulty of project management,the reason lies in the dual constraints of resources and processes. To solve the project schedule is an NP-hard problem under the constraints, especially in multi-resource constraints adds to the problem complexity. It adopts critical chain theory and method to analyze the problem and construct the model. Apply adaptive genetic al- gorithm to solve the problem. Genetic coding, selection, crossover and mutation are discussed, and on this basis, the traditional genetic al- gorithm is improved. Finally, based on the experiment analysis of the instance, verifiy the validity of the critical chain method in the study of multi-resource project scheduling problem, as well as the feasibility of adaptive genetic algorithm to solve the problem.

关 键 词：项目进度计划 多资源约束 关键链 遗传算法 自适应遗传算法 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]