摘要
针对局部地形复杂、振荡强烈的函数优化精度难以提高的问题,提出一种自动调整邻域搜索范围和方向的自适应变邻域混沌搜索微粒群算法(AVNC-PSO)。优化初期首先由基本PSO算法进行粗调,当种群收敛于局部最优时,选择飞行停滞且聚集程度高的粒子向不同方向的邻域内进行混沌搜索,搜索方向和粒子偏移量根据粒子与收敛中心的距离和混沌变量的值共同确定。数值仿真表明,该算法能够使局部搜索更精确,有效改善基本PSO算法优化精度不高的弱点。
An adaptive variable neighborhood chaos search PSO that can automatically adjust neighborhood range and direction is proposed for optimization of functions,which is with complex terrain and strong oscillation.Firstly optimize with standard PSO, when the particle swarm converges at local best solution,choose the particles that are stagnant and high extent convergence,and make them chaos search towards various directions in the neighborhood.Both the distance between particle and the convergence center and the chaos variable determine the direction and particle position offset.Simulation results show that this method makes higher precise of local optimization and can improve the algorithm performance effectively.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第31期90-92,共3页
Computer Engineering and Applications
基金
浙江省教育厅2006年度高校科研计划(No.20060347)
关键词
自适应
变邻域
混沌搜索
微粒群算法
self-adaptive
variable neighborhood
chaos search
Particle Swarm Optimization（PSO）
