电子学报 ›› 2014, Vol. 42 ›› Issue (9): 1831-1838.DOI: 10.3969/j.issn.0372-2112.2014.09.026

• 科研通信 • 上一篇    下一篇

多策略差分进化的元胞多目标粒子群算法

朱大林, 詹腾, 张屹, 郑小东   

  1. 三峡大学水电机械设备设计与维护湖北省重点实验室, 湖北宜昌 443002
  • 收稿日期:2013-05-28 修回日期:2013-10-03 出版日期:2014-09-25
    • 通讯作者:
    • 张 屹
    • 作者简介:
    • 朱大林 男,1957年10月生于湖北宜昌,三峡大学机械与动力学院教授,主要研究方向:系统优化设计、机械及结构可靠性. E-mail:dlzhu@ctgu.edu.cn;詹 腾 男,1988年2月生于湖北黄冈,三峡大学硕士研究生,主要研究方向:智能优化算法,工程优化设计等. E-mail:zhanteng1988@gmail.com;郑小东 男,1989年6月生于湖北孝感,三峡大学硕士研究生,主要研究方向智能优化算法.
    • 基金资助:
    • 国家自然科学基金 (No.51275274)

Cellular Multi-Objective Particle Swarm Algorithm Based on Multi-Strategy Differential Evolution

ZHU Da-lin, ZHAN Teng, ZHANG Yi, ZHENG Xiao-dong   

  1. Hubei Key Laboratory of Hydroelectric Machinery Design & Maintenance, China Three Gorges University, Yichang, Hubei 443002, China
  • Received:2013-05-28 Revised:2013-10-03 Online:2014-09-25 Published:2014-09-25
    • Supported by:
    • National Natural Science Foundation of China (No.51275274)

摘要:

为了增加Pareto解集的多样性,平衡多目标优化的全局搜索和局部寻优的能力,提出一种多策略差分进化的元胞多目标粒子群算法.该算法在分析粒子群优化原理基础上,将元胞自动机理论融入粒子群算法,研究粒子种群的交流结构和信息传递机制.为了避免粒子飞行速度过快陷入局部收敛,提出一种限制粒子飞行速度的策略,并引入一种多策略差分进化选择算子增加对粒子的扰动.实验证明,该算法相对于比较算法,有更好的收敛性和多样性.

关键词: 元胞自动机, 粒子群算法, 多策略差分进化, 速度控制策略, 多目标优化

Abstract:

In order to strengthen the diversity of Pareto sets obtained by multi-objective optimization algorithms and balance the exploration and exploitation of the algorithms,a cellular multi-objective particle swarm optimization algorithm based on multi-strategy differential evolution (MPSOCell) is proposed.This algorithm is composed by integrating the cellular automate mechanism into the basic particle swarm optimization algorithm,and it is aimed at promoting the communication and information transmission among the particles.To avoid the local convergence caused by the fast flying speed of particles,a strategy used to limit the flying speed is designed;to strengthen the disturbance to the particles,a multi-strategy differential evolution operator is also brought into the algorithm.The experiments demonstrate that MPSOCell has better performance in terms of convergence and diversity.

Key words: cellular automata, particle swarm optimization, multi-strategy differential evolution, speed control strategy, multi-objective optimization

中图分类号: