Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (18): 48-50.DOI: 10.3778/j.issn.1002-8331.2009.18.015

• 研究、探讨 • Previous Articles     Next Articles

Learning-exam particle swarm optimization algorithm

DAI Jun,LI Guo,XU Chen   

  1. Institute of Intelligent Computing Science,College of Mathematics and Computational Science,Shenzhen University,Shenzhen,Guangdong 518060,China
  • Received:2009-01-15 Revised:2009-03-23 Online:2009-06-21 Published:2009-06-21
  • Contact: DAI Jun

学习-考试型的粒子群优化算法

代 军,李 国,徐 晨   

  1. 深圳大学 数学与计算科学学院 智能计算科学研究所,广东 深圳 518060
  • 通讯作者: 代 军

Abstract: The Standard Particle Swarm Optimization(SPSO) algorithm usually sinks into the local optimal search space at the later stage of the particles’ evolution.To improve the ability of searching global optimal value of particle swarm optimization algorithm,firstly the basic principle of learning-exam mechanism is gotten based on the analysis of students’ learning-exam mechanism.When particle sinks into the local optimal space,its position vector components is compounded organically based on the principle and the particle’s local optimization,that’s the exam strategy.Then numeric experiments indicate that the new strategy greatly improves the particles’ ability of searching global optimal value.

Key words: Particle Swarm Optimization(PSO), evolutionary computing, strategy, learning-exam

摘要: 标准的粒子群算法在进化后期常易于陷入局部最优。为提高粒子群算法的寻优性能,首先对学生学习-考试机制进行分析,得到学习-考试机制的基本原则,然后,利用该原则和粒子局部最优的信息,在粒子陷入局部最优时,对粒子的位置分量进行有机地组合,即考试策略。数值实验结果证明了新策略极大地提高了粒子的寻优性能。

关键词: 粒子群优化, 进化计算, 策略, 学习-考试