计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (5): 39-42.

• 理论研究、研发设计 • 上一篇    下一篇

基于多点力学的粒子群算法改进

沈佳杰,江  红,王  肃   

  1. 华东师范大学 信息科学技术学院,上海 200241
  • 出版日期:2014-03-01 发布日期:2015-05-12

Particle Swarm Optimization based on multi-point mechanics

SHEN Jiajie, JIANG Hong, WANG Su   

  1. School of Information Science and Technology, East China Normal University, Shanghai 200241, China
  • Online:2014-03-01 Published:2015-05-12

摘要: 针对于粒子群算法(PSO)易早熟、参数相关性强以及高维数据难以优化的问题。通过引入多点力学计算和理论建模推导,提出了一个新的粒子速度的计算方法。通过理论推导和实验证明对理论结果进行仿真和测试,对于常见的测试函数,新粒子群算法较标准粒子群算法有良好的适应性和准确性的改进。

关键词: 粒子群算法, 化计算, 随机搜索

Abstract: To deal with the problems of the prematureness, strong pertinence of the parameters, and high-dimensional data in Particle Swarm Optimization(PSO), a new algorithm to calculate the velocity of the particles is proposed. The better adaptability and accuracy of the new algorithm is verified by theoretical derivation and experimental test.

Key words: Particle Swarm Optimization, evolutionary computation, stochastic searching