基于动态概率变异的Cauchy粒子群优化  

Cauchy Particle Swarm optimization based on dynamic probability mutation

作　　者：刘栋[1] 郝婷[2] 刘希玉[1] 

机构地区：[1]山东师范大学信息科学与工程学院,济南250014 [2]山东大学材料科学与工程学院,济南250061

出　　处：《计算机工程与应用》2007年第16期77-79,共3页Computer Engineering and Applications

基　　金：山东省自然科学基金(the Natural Science Foundation of Shandong Province of China under Grant No.Z2004G02);山东省教育厅资助科研课题(the research Project of Department of Education of Shandong Province;China under Grant No.J05G01);"泰山学者"建设工程专项经费资助

摘　　要：介绍了标准粒子群优化(SPSO)算法,在两种粒子群改进算法Gaussian Swarm和FuzzyPSO的基础上提出了Cauchy粒子群优化(CPSO)算法,并将遗传算法中的变异操作引入粒子群优化,形成了动态概率变异Cauchy粒子群优化(DMCPSO)算法。用3个基准函数进行实验,结果表明,DMCPSO算法性能优于SPSO和CPSO算法。This paper first introduce Standard Particle Swarm Optimization algorithm,and then Cauchy Particle Swarm Optimization,which is based on two improved particle swarm optimization algorithms：Gaussian Swarm and Fuzzy PSO,is proposed.Moreover,it imports mutation operation of Genetic Algorithm to Particle Swarm Optimization and form dynamic probability mutation Cauchy Particle Swarm Optimization algorithm.Three benchmark function are tested and show that the performance of the DMCPSO Algorithm is better than SPSO and CPSO algorithms.

关 键 词：粒子群优化 GAUSSIAN SWARM FUZZY PSO 变异 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]