基于KS函数求解多约束优化问题研究  

Study on solving problems of multi-constraint optimization based on KS function

作　　者：万朝燕[1] 兆文忠[1] 李晓峰[1] 马思群[1] 

机构地区：[1]大连铁道学院机械工程系,辽宁大连116028

出　　处：《机械设计》2003年第9期9-11,共3页Journal of Machine Design

基　　金：国家"八六三"高技术研究发展计划资助项目 (2 0 0 1AA41 1 1 1 0 )

摘　　要：针对机械优化设计中常见的多约束优化问题 ,提出了一种求解新方法 :利用KS函数的凝聚特性 ,将优化问题中的全部或部分约束凝聚为一个近似的、逼近精度仅由一个参数控制的约束 ,然后 ,再对约束压缩后的优化问题选用适当的约束优化方法求解。三维空间中一例可行域的图形给出了约束凝聚的几何解释。实例证明 ,由于压缩了求解规模 。Aimed at the common problems of multi-constraint optimization in optimal design of machinery,a new kind of solving method was presented.Utilizing the agglomeration features of KS function,the total or partial constraints of optimization problems were agglomerated to an approximate and draw near precision,a constraint to be controlled by one parameter only.And then choose proper constraint optimization method to solve the optimal problem after the compression of constraints.The geometric explaination of constraints agglomeration was presented by an example of feasible region figure in three dimensional space.

关 键 词：KS函数 可行域 凝聚算法 多约束优化问题 机械设计 优化设计 

分 类 号：TH122[机械工程—机械设计及理论] O224[理学—运筹学与控制论]