复杂机械产品性能意图优化研究  

Performance intention optimization for complex mechanical products

作　　者：高一聪[1] 谭建荣[1] 冯毅雄[1] 魏喆[1] 

机构地区：[1]浙江大学CAD&CG国家重点实验室,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2008年第9期1549-1553,1596,共6页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(50835008;50875237);国家"863"高技术研究发展计划资助项目(2007AA04Z190);国家科技支撑计划资助项目(2006BAF01A37)

摘　　要：为解决专用装备制造业中复杂机械产品多设计参数、多设计约束、多设计目标的设计优化问题,采用一种基于Pareto最优解的多目标广义差分进化(GDE)算法.GDE算法在普通差分进化算法的基础上改进了约束条件的处理方法,利用差分进化算法模拟生物种群进化,以群体中的所有个体为对象,采用随机化技术对一个被编码的参数空间进行高效搜索,通过反复迭代搜索到多目标优化问题的近似Pareto最优解集.以高速工业平缝机勾线机构的设计为例,通过对求得的Pareto最优解集的分析,得到了高速工业平缝机勾线机构的设计特性,为设计人员今后进行高速工业平缝机勾线机构性能设计优化提供了充分的依据.A multi-objective generalized differential evolution algorithm based on Pareto optimal solution was adopted to solve the multi-parameter, multi-constraint and multi-objective optimization problems of complex mechanical products. The generalized differential evolution algorithm uses an improved constraints handling approach based on the differential evolution algorithm. It simulates the process of evolutionary to get the Pareto front by searching the coded parameter space efficiently based on randomiza- tion techniques. A thread hooking mechanism of high-speed lockstitch sewing machine was studied as an example. According to the analysis of the obtained solutions, the design characteristics and principles of the thread hooking mechanism of high-speed lockstitch sewing machine were concluded, which is otherwise difficult to obtain using other means.

关 键 词：勾线机构 高速工业平缝机 广义差分进化算法 性能优化 

分 类 号：TH122[机械工程—机械设计及理论]