您的位置:首页  > 论文页面

粒子群优化算法变型:GLbest-PSO算法

发表时间:2016-01-15  浏览量:2414  下载量:1053
全部作者: 陈相托,王惠文,孙玥元
作者单位: 北京航空航天大学中法工程师学院;巴黎中央理工学院;北京航空航天大学经济管理学院
摘 要: 全局粒子群优化(global particles swarm optimization,Gbest-PSO)算法拥有较快的收敛速度,但粒子群体容易在局部过早收敛;而局部粒子群优化(local particles swarm optimization,Lbest-PSO)算法细化了粒子群对空间的探索,但是弱化了其最优解的聚拢效应。因此,研究尝试中和Gbest-PSO算法和Lbest-PSO算法在全局搜索和局部搜索方面各自的优点,提出GLbest-PSO算法模型,并希望以此获得一种兼有收敛速度和搜索细化的优化算法。通过仿真实验检验该算法的优化能力,并与其他几种经典的粒子群算法进行比较。结果表明,该算法在收敛速度和收敛效果上均有所提高,是一种有效的改进算法。
关 键 词: 最优化;粒子群优化;全局
Title: A new variant of particle swarm optimization method: the GLbest-PSO
Author: CHEN Xiangtuo, WANG Huiwen, SUN Yueyuan
Organization: Ecole Centrale de Pekin, Beihang University; Ecole Centrale Paris; School of Economics and Management, Beihang University
Abstract: Global particles swarm optimization (PSO), noted as Gbest-PSO, has a faster convergence, but it causes easily the premature and local convergence. However, the local PSO algorithm, noted as Lbest-PSO, makes the research in space more refined but weakens the effect of the optimal particles. In this paper, trying to neutralize the gathering effects of Gbest-PSO and Lbest-PSO, we have proposed a GLbest model in order to have an optimization algorithm, which is able to obtain both a fast convergence and a refined search in space. GLbest-PSO and some other variants are tested on a common used set of optimization functions and their capacities are compared in the end. The results show that the convergence speed and convergence effect of the propsosed algorithm are improved, which proved to be an effective method.
Key words: optimization; particles swarm optimization; global-local; topology neighborhood
发表期数: 2016年1月第1期
引用格式: 陈相托,王惠文,孙玥元. 粒子群优化算法变型:GLbest-PSO算法[J]. 中国科技论文在线精品论文,2016,9(1):75-80.
 
1 评论数 0
暂无评论
友情链接