基于遗传算法的机电产品布线结构优化设计方法  

Configuration optimization for electromechanical products cable harness design based on genetic algorithm

作　　者：李国闻 张丹[1] 杜海遥 左敦稳[1] 

机构地区：[1]南京航空航天大学机电学院,南京210016

出　　处：《中国科技论文》2015年第16期1944-1948,1952,共6页China Sciencepaper

基　　金：高等学校博士学科点专项科研基金资助项目(20133218120032);航空科学基金项目(20141652016);南京航空航天大学基本科研业务费资助项目(NS2013047)

摘　　要：机电产品中线缆的捆扎结构是布线路径设计的基础。针对目前布线结构的设计严重依靠经验反复尝试,结构设计缺乏优化的问题,本文提出了一种基于改进遗传算法的机电产品线缆布线结构优化设计方法。首先对需要布线的机电产品布线环境进行了预处理,然后针对常规遗传算法在计算线缆拓扑结构的过程中致死率高和容易陷入局部最优解等的问题,提出了线缆结构的初始化算法,改进了变异算法和杂交算法,通过对致死基因进行定向变异,提高了常规遗传算法中随机变异的成功率。通过实例显示,改进后的算法可获得优化的线缆结构,所提算法显著降低了致死率,具有较好的收敛性。Binding structure of cable in mechatronic products is the basis of wiring path design.Considering the fact that the cable structure design heavily relies on the experience of trial and the cable structure design is lack of optimization,this paper proposed an improved genetic algorithm to calculate the cable structure.This paper treated wiring environment for the pretreatment of the mechatronic products at first,and a new algorithm for initialization was then proposed in consideration of the high fatality rate of the traditional genetic algorithm in the process of calculating cable topology and its result is more likely to fall into local optimal solution.This paper also improved the mutation algorithm and hybrid algorithm,and improved the success rate of the random variation in the traditional genetic algorithm by directional variation of lethal gene mutation.The examples show that the im-proved algorithm can obtain the optimized cable structure,significantly reduce the fatality rate,and have a good convergence.

关 键 词：遗传算法 布线设计 拓扑结构优化 机电产品 

分 类 号：TB492[一般工业技术] TP18[自动化与计算机技术—控制理论与控制工程]