计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (2): 44-47.DOI: 10.3778/j.issn.1002-8331.2009.02.012

• 研究、探讨 • 上一篇    下一篇

粒子群算法的新改进思想—递进式搜索

权 文,王晓丹,王 坚,俞 群   

  1. 空军工程大学 导弹学院,陕西 三原 713800
  • 收稿日期:2008-07-22 修回日期:2008-09-27 出版日期:2009-01-11 发布日期:2009-01-11
  • 通讯作者: 权 文

Improved particle swarm optimization algorithm:increase by degrees search

QUAN Wen,WANG Xiao-dan,WANG Jian,YU Qun   

  1. Missile Institute,Air Force Engineering University,Sanyuan,Shaanxi 713800,China
  • Received:2008-07-22 Revised:2008-09-27 Online:2009-01-11 Published:2009-01-11
  • Contact: QUAN Wen

摘要: 针对传统粒子群算法的不足,提出了一种改进思想:分“初选”和“细搜”两个阶段搜索,分别设立评价能力递增的不同适应度函数和采用不同的惯性权重,进行递进式搜索;构造了一种新的网格构造图——中心对称标距图,并结合目前广泛研究的航迹规划问题,设定3种不同情况对改进算法进行了仿真验证。结果表明,改进后的粒子群算法寻优能力增强,收敛速度也有所提高。

关键词: 粒子群算法, 中心对称标距图, 优化, 航迹规划

Abstract: Aimed at the shortcomings of traditional Particle Swarm Optimization(PSO),an improved algorithm is proposed.The search is divided into two parts:selection in the beginning and fine-draw search subsequently.In order to search by degrees,different increased evaluation adaptability functions and different inertia weighs are set.The central symmetry marked drawing is proposed,which is a new grid constructional drawing.Integrate with study of flight route planning,the improved algorithm is tested by three different situation and shows that the new algorithm is better than the traditional particle swarm optimization with both a better search capacity and a quickly speed of convergence.

Key words: particle swarm optimization, central symmetry marked drawing, optimization, route planning