北京亚运村汽车交易新市场工程钢结构的质量控制  

An Improved Genetic Algorithm and Its Application to Risk Multi-Objective Decision-making in Construction Project

作　　者：李丹[1] 袁永博[1] 张明媛[1] LI Dan, YUAN Yong-bo, ZHANG Ming-yuan （ School of Civil and Hydraulic Engineering, Dalian University of Technology, Dalian 116024, China ）

机构地区：[1]大连理工大学土木水利学院,辽宁大连116024

出　　处：《建筑管理现代化》2008年第5期79-82,共4页Construction Management Modernization

摘　　要：针对多目标遗传算法NSGA—Ⅱ的缺点，提出了一种改进算法。将改进的遗传算法应用到工程风险的多目标决策中，通过实例分析证明了其在解决工程项目风险管理多目标优化问题上有效性。将改进前后两种算法进行对比分析，证明了改进的遗传算法在保证Pareto最优边界收敛性的同时改善了Pareto最优解的多样性。An improved genetic algorithm is proposed to overcome the disadvantages of NSGA-Ⅱ. The improved genetic algorithm is applied to risk multi-objective decision-making in construction project. An application example is analyzed to demonstrate the efficiency of the new algorithm in solving multi-objective problems of construction risk management. The result also shows that the new method is capable of improving convergence and variety of Pareto front when compared with ordinary NSGA-Ⅱ.

关 键 词：工程项目 风险决策 多目标 遗传算法 

分 类 号：F282[经济管理—国民经济]