嵌入不动点迭代法的粒子群算法  

Particle Swarm Optimization Algorithm Based on Fixed-point Iteration Method

作　　者：刘微[1] 王洪瑞[1] 陈爽[1] 

机构地区：[1]吉林师范大学信息技术学院,吉林四平136000

出　　处：《吉林师范大学学报（自然科学版）》2017年第2期80-86,共7页Journal of Jilin Normal University:Natural Science Edition

基　　金：国家自然科学基金项目(61305082)

摘　　要：针对利用粒子群优化算法寻找最优极值的问题,本文提出了一种对粒子群优化算法的改进方式.在粒子群优化算法中,加入不动点迭代法,即在进行粒子群迭代寻优之后,再利用不动点迭代策略再次进行迭代寻优计算.通过两次迭代寻优计算,让算法更加快速的收敛到最优值,并且能够找到更小的极值,且不易陷入局部极值.让粒子群算法更加稳定,寻优更加优化.This paper proposes a noval method to improve the Particle Swarm Optimization algorithm in finding the optimal extreme value problem. We add the Fixed-point iterative method to the Particle Swarm Optimization algorithm, and proposed a new optimization algorithm, which is called F-PSO. The idea of F-PSO is that when the iterative optimization of the Particle Swarm Optimization is finished ,the iterative optimization calculation is carried out again by using the Fixed Point Iteration strategy. Through the two iterative optimization calculations ,the algorithm can converge to the optimal value more quickly, and can find the smaller extreme value as well. In addition, F-PSO is not easy to fall into local extremum, it is more stable and more optimized than Particle Swarm Optimization algorithm.

关 键 词：粒子群优化算法 不动点迭代法 寻优计算 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]