混沌映射的粒子群优化方法  

Method of particle swarm optimization based on the chaos map

作　　者：刘道华[1,2] 原思聪[2] 兰洋[1] 马新建[2] 

机构地区：[1]信阳师范学院计算机与信息技术学院,河南信阳464000 [2]西安建筑科技大学机电工程学院,陕西西安710055

出　　处：《西安电子科技大学学报》2010年第4期764-769,共6页Journal of Xidian University

基　　金：河南省自然科学基金资助项目(2010070018);河南省教育厅自然科学研究计划资助项目(2010A520034);河南省高等学校青年骨干老师计划资助项目(2009A088)

摘　　要：为提高粒子群优化的求解性能,在分析了粒子群优化原理的基础上,给出了两种混沌映射的映射规则.构建了基于Logistic映射的混沌粒子群优化方法以及基于Lozi s映射的混沌粒子群优化方法,并给出了两类约束条件的处理方法.采用基于Logistic映射的混沌粒子群优化方法和基于Lozi s映射的混沌粒子群优化方法以及标准粒子群优化方法分别对benchmark有约束优化实例进行求解.对各种方法获得的最优解、成功率指标、平均有效迭代数、迭代占用时间等方面作对比,结果表明:采用基于Lozi s射映的混沌粒子群优化方法具有求解精度高、优化效率高等优点.In order to improve the solving performance of particle swarm optimization（PSO）,first the basic principle of PSO is analyzed and mapping rules for two types of chaos map are given.Next,PSO based on the Logistic map（LGM-PSO） and Lozi＇s map（LZM-PSO） are constructed,and the methods for treating two types of constraints are given.To compare the performances of LGM-PSO,LZM-PSO and the standard particle swarm optimization,the three methods are used to solve the benchmark constrained optimal problem.Their performances are compared in terms of optimal solution,success ratio,average valid evaluation number,iterative occupancy hours and so on.Comparison results indicate that the LZM-PSO has many advantages,such as higher solution accuracy and higher computational efficiency.

关 键 词：Logisitic映射 Lozis映射 混沌理论 粒子群优化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]