混沌_微粒群算法及其在机械优化设计中的应用.pdf

23 5 Vo.l 23No. 5 2007 10 M achine Design and Research Oct. , 2007 文章编号: 2007) 0500603 , (福州大学 机械工程与自动化学院, 福建 福州 350002, Emai:l cnsm liuh@ tom. com) : 针对机械优化设计中变量多目标函数和约束条件复杂而难以求解的问题, 立了基于微粒群算 法的机械优化设计的数学模型;并针对传统罚函数法处理约束条件而引起的病态问题,提出一种利用混沌变量 来更新产生违约解 个体的方式来改进微粒群算法, 增加了 个体的多样性避免微粒群算法出现早熟, 从而加快 算法的收敛速度实例计算表明该算法能较好地解决机械优化设计问题 : ;; ; : TH 122 : ChaosParticle Swarm Optim ization Algorithm and Its Application toM echanicalOptim alDesign LIU Hong, ZH NG QiShan ( Sch. ofM echanicalEng. utomation, Fuzhou Univ, Fuzhou 350002, China) Abstract: mi ed at the problems ofmultivariable and complicated target function and constraint conditions inme chanical optmi al design, themathematicalmodel combing mechanical optmi al designw ith Particle Swarm Optmi ization ( PSO) is constructed. Due to the morbidity problem of traditional penalty function dealing w ith the constraint condi tions inmechanicaloptmi aldesign, themethod using Chaosvariable to update the individualparticle producing infeasi ble solution is proposed for mi proving PSO. The mi proved ChaosPSO algonthm can increase individual diversity, a void prematurity ofPSO and speed up the step of convergence. The practical example show s that the result ofmechani cal optmi al design based on ChaosPSO is remarkable. Key words: chaos; particle swarm optmi ization; optmi aldesign; constraint conditions ( Particle Swarm Optmi ization, PSO)[ 1~ 3] : J Kennedy R C Eberhart 1995 m inf (x), x Rn ( 1) , ,