计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (31): 42-44.

• 学术探讨 • 上一篇    下一篇

具有综合学习机制的粒子群算法

唐岑琦,周育人

  

  1. 华南理工大学 计算机工程与科学学院,广州 510640
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-11-01 发布日期:2007-11-01
  • 通讯作者: 唐岑琦

Particle swarm optimization algorithm with synthesis study mechanism

TANG Cen-qi,ZHOU Yu-ren   

  1. Academy of Computer Science and Engineering,South China University of Technology,Guangzhou 510640,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-11-01 Published:2007-11-01
  • Contact: TANG Cen-qi

摘要: 基本粒子群算法在模拟生物群体智能时,只有信息的单一传递和强迫学习机制,导致群体迅速收敛和种群的多样性降低。为此,提出一种具有综合学习机制的粒子群算法,将所有粒子的个体极值的平均值取代每一粒子的个体极值,并以自适应概率定向地随机变异全局极值。仿真实验表明,新算法解精度高,收敛速度快,能有效抑制过早收敛。

Abstract: The basic particle swarm optimization algorithm,when simulating biology community intelligence,only has the sole transmission of information and force study mechanism,it causes the swarm rapidly to restrain and population’s diversity reduces.Therefore,this paper proposed a particle swarm optimization algorithm with synthesis study mechanism.It took the meanvalue of all particles’pbest as each particle’pbest,and varied stochastically the gbest by auto-adapted probability in the direction detection.The simulation experiment indicated that,new algorithm had a high solution precision and a quick convergence rate,and it could effectively suppress premature restrains.