一种改进的混沌粒子群优化混合算法.pdfVIP

第39 卷第1 期 应 用 科 技 Vol.39, No.1 2012 年2 月 Applied Science and Technology Feb. 2012 doi:10.3969/j.issn.1009-671X.201111004 一种改进的混沌粒子群优化混合算法 钱晓山1, 2 1. 宜春学院 物理科学与工程技术学院，江西 宜春, 336000 2. 中南大学 信息科学与工程学院，湖南 长沙, 410083 摘 要：提出了一种改进的混沌粒子群优化混合算法. 该算法利用信息交换机制将两组种群分别用差分进化算法和 粒子群算法进行协同进化，并且将混沌变异操作引入其中，加强算法的局部搜索能力. 通过对 3 个标准函数进行测 试，仿真结果表明该算法与差分进化粒子群优化（DEPSO ）算法相比，全局搜索能力和抗早熟收敛性能大大提高. 关键词：混合算法；差分进化；粒子群优化；协同进化；混沌变异；早熟收敛 中图分类号：TP18 文献标志码：A 文章编号：1009-671X （2012 ）01-0005-04 A hybrid algorithm of the improved chaotic particle swarm optimization QIAN Xiaoshan1,2 1. College of Physical Science and Engineering Technology, Yichun University, Yichun 336000, China 2. School of Information ScienceEngineering, Central South University, Changsha 410083, China Abstract: A hybrid algorithm of the improved chaotic particle swarm optimization is proposed. Based on the information exchange mechanism, the algorithm uses differential evolution algorithm and particle swarm algorithm to make co-evolution for two groups of populations, and the chaos mutation is introduced into the algorithm to enhance the efficiency of local search capabilities. Using three standard functions to test it, simulation results show that, compared with the differential evolution and particle swarm optimization （DEPSO ）algorithm, global search ability and resistance to premature convergence are increased greatly. Keywords: hybrid algorithm; differential evolution optimization; particle swarm; co-evolution; chaotic variation; premature convergence 1由 Price[1]首先提出的一种基于