作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2013, Vol. 39 ›› Issue (5): 200-203,208. doi: 10.3969/j.issn.1000-3428.2013.05.044

• 人工智能及识别技术 • 上一篇    下一篇

基于多种群多模型协同进化的粒子群优化算法

徐冰纯,葛洪伟,王燕燕   

  1. (江南大学物联网工程学院,江苏 无锡 214122)
  • 收稿日期:2012-05-31 出版日期:2013-05-15 发布日期:2013-05-14
  • 作者简介:徐冰纯(1988-),女,硕士研究生,主研方向:人工智能;葛洪伟,教授、博士生导师;王燕燕,硕士研究生

Particle Swarm Optimization Algorithm Based on Multi-swarm and Multi-model Cooperative Evolution

XU Bing-chun, GE Hong-wei, WANG Yan-yan   

  1. (School of Internet of Things Engineering, Jiangnan University, Wuxi 214122, China)
  • Received:2012-05-31 Online:2013-05-15 Published:2013-05-14

摘要: 为克服标准粒子群优化(PSO)算法易陷入局部极值和优化精度较低的缺点,提出一种多种群多模型协同进化的粒子群优化(MSM-PSO)算法。将整个粒子群分成大小相等的3个分群,各分群采用不同的进化模型,分群间相互影响促进。同时采用自适应动态惯性权重,以保持种群多样性,降低陷入局部极值的概率。测试结果表明,该算法全局性能好、寻优精度高。

关键词: 粒子群优化算法, 多种群, 多模型, 自适应动态惯性权重, 协同进化

Abstract: Aiming to improve the performance of standard Particle Swarm Optimization(PSO) algorithm, an improved PSO algorithm based on Multi-swarm and Multi-model cooperative evolution(MSM-PSO) is proposed. The particles are divided into three swarms in the same size, different swarms adapt different evolution models, and three swarms interact and affect each other. Besides, adaptive dynamic inertia weight is adopted in the algorithm, which can maintain diversity of population and reduce the possibility of local minimum. Simulation result shows that this method has better global performance, and higher optimization precision.

Key words: Particle Swarm Optimization(PSO) algorithm, multi-swarm, multi-model, adaptive dynamic inertia weight, cooperative evolution

中图分类号: