改进遗传算法用于给水管网多工况优化设计  

Improved genetic algorithm for optimal design of water supply network in variable regime

作　　者：王瑛[1] 李文娟[1] 李梦[1] 

机构地区：[1]兰州理工大学土木工程学院,甘肃兰州730050

出　　处：《兰州理工大学学报》2011年第3期129-132,共4页Journal of Lanzhou University of Technology

基　　金：甘肃省自然科学基金(3ZS042-B25-005)

摘　　要：为满足给水管网优化设计中多工况约束条件,达到全局优化的范畴,提出以给水管网年折算费用最低的多工况设计数学模型,并从适应度值标定和群体多样化两方面来改进遗传算法,应用该遗传算法求解模型.对工程算例管网进行优化求解,优化结果的对比表明,采用改进遗传算法能在满足管网运行各个工况的约束条件下,年折算费用比采用常规算法节省2.63%.In order to satisfy the variable regime constraint in optimization of pipe network design and achieve global optimization,a mathematical model for designing water supply network in variable regime was presented with minimum conversion cost and the genetic algorithm was improved in terms of both the fitness value and group diversification.This improved genetic algorithm was then used to solve the model.Finally,an engineering example was given to show the optimization procedure.It was shown by contrasting the optimization results that using improved genetic algorithm could meet foregoing constraint conditions of variable regime of the network and save 2.63% of conversion cost when compared to using conventional algorithm.

关 键 词：给水管网 优化设计 多工况 遗传算法 适应度值 群体多样化 

分 类 号：TU991[建筑科学—市政工程]