自适应变异的离散粒子群算法研究  

A Study on the Particle Swarm with Adaptive Mutation

作　　者：张元清[1] 包骏杰[1] 卫茜[2] 

机构地区：[1]重庆教育学院计算机与现代教育技术系,重庆400067 [2]重庆大学外国语学院,重庆400044

出　　处：《重庆三峡学院学报》2007年第3期27-30,共4页Journal of Chongqing Three Gorges University

摘　　要：针对算法易早熟和收敛慢的缺点,提出了一种新的自适应变异离散粒子群算法。算法中的变异思想是一种确定性交异操作,能使算法中陷入局部极小区域的粒子通过变异行为进行全局寻优,从而克服算法易早熟的缺陷。To overcome the weakness that the arithmetic at present is prone to premature and is slow to convergence,a completely new way is found out,that is, SAMDPSO. The idea of this arithmetic is a definite modulating operation, which enables the particles trapped in local tiny field to search for optimization roundly through modulation. Thus, disadvantage of prematudty in computing was overcome.

关 键 词：离散粒子群算法 自适应变异 TSP在问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]