电子学报 ›› 2013, Vol. 41 ›› Issue (5): 865-870.DOI: 10.3969/j.issn.0372-2112.2013.05.006

• 学术论文 • 上一篇    下一篇

一种基于动态边界的粒子群优化算法

李迎秋1,2, 迟玉红3, 温涛1,2   

  1. 1. 东北大学软件中心,辽宁沈阳 110004;
    2. 大连东软信息学院计算机系,辽宁大连116023;
    3. 中国人民解放军65053部队,辽宁大连116113
  • 收稿日期:2012-03-16 修回日期:2012-10-10 出版日期:2013-05-25
    • 作者简介:
    • 李迎秋 女,1972年出生于河北,东北大学软件中心博士研究生,大连东软信息学院计算机系教师,副教授.研究方向为计算智能、服务计算. E-mail:liyingqiu@neusoft.edu.cn
    • 基金资助:
    • 国家自然科学基金 (No.61170168,No.61170169)

A Dynamic Boundary Based Particle Swarm Optimization

LI Ying-qiu1,2, CHI Yu-hong3, WEN Tao1,2   

  1. 1. Software Center,Northeastern University,Shenyang,Liaoning 110004,China;
    2. Department of Computer Science and Technology,Dalian Neusoft Information Institute,Dalian,Liaoning 116023,China;
    3. Unit 65053,PLA,Dalian,Liaoning 116113,China
  • Received:2012-03-16 Revised:2012-10-10 Online:2013-05-25 Published:2013-05-25

摘要: 2007年提出的标准粒子群优化算法(PSO-2007)在进化的后期容易出现停滞现象而导致早熟收敛,为此本文提出了一种基于动态边界的粒子群优化算法(DBPSO).该算法根据停滞期粒子运动的特点,将边界动态调整策略引入到PSO-2007中,通过跟踪粒子飞行位置的分布动态调整搜索空间的边界,引导粒子在更有效的区域内进行搜索,从而减轻早熟收敛,提高收敛精度.典型测试函数的求解实验结果表明DBPSO是可行而有效的.

关键词: 粒子群优化, 停滞现象, 早熟收敛, 动态边界

Abstract: Standard particle swarm optimization presented in 2007(namely,PSO-2007)inclines towards stagnation phenomena in the later stage of evolution,which leads to premature convergence.Therefore,a PSO based on dynamic boundary(namely,DBPSO)is proposed in this paper.According to the movement characteristics of particles at stagnation stage,DBPSO introduces a strategy of boundary adjusting in PSO-2007.By tracking the distribution of the particles’locations,DBPSO adjusts the boundary of search space dynamically,which could guide the particles to more promising region.This strategy helps PSO-2007 decrease premature convergence and improve convergence precision.The results of experiments of four typical functions show that DBPSO are feasible and effective.

Key words: particle swarm optimization, stagnation phenomena, premature convergence, dynamic boundary

中图分类号: