合作型协同多目标群搜索算法  

Cooperative Coevolutionary Multi-objective Group Search Optimizer

作　　者：李亚洲[1,2] 郑向伟[1,2] 肖宪翠 

机构地区：[1]山东师范大学信息科学与工程学院,济南250014 [2]山东省分布式计算机软件新技术重点实验室,济南250014

出　　处：《小型微型计算机系统》2016年第3期567-571,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61373149;61272094;61472232)资助

摘　　要：群搜索算法(Group Search Optimizer,GSO)是一种基于动物群体行为的智能优化算法,在高维函数优化和收敛性方面表现出良好性能.本文基于分而治之策略和协同进化框架,提出了一种合作型协同多目标群搜索算法(Cooperative Coevolutionary M ulti-Objective GSO,CM OGSO).首先将群(group)划分为多个子群(sub-groups),采用改进的群搜索算法演化每个子群,其次选择其它子群中处于非支配位置的成员(member),构建当前子群的成员的上下文向量,通过目标函数评价子群成员.最后,结合各个子群的成员构建多目标问题的Pareto解集.实验结果表明,相比于其他多目标优化算法,CMOGSO算法所求Pareto解集具有精度高、解分布均匀等优势,能够有效地解决多目标优化问题.Group Search Optimizer（ GSO） is a swarm intelligence algorithm inspired from animal＇s foraging behavior. Its superiority is demonstrated in high dimensional function optimization. Based on the strategy of divide-and-conquer and cooperative coevolution framework,a Cooperative Coevolutionary Multi-objective Group Search Optimizer（ CMOGSO） is proposed in this paper. In CMOGSO,multiobjective optimization problems are decomposed according to their decision variables and are optimized by improved GSO respectively.Collaborators are selected randomly from archive and employed to construct context vectors in order to evaluate the members in subgroups. Experimental results demonstrate that CMOGSO can more effectively and efficiently solve multi-objective optimization problemsand the accuracy and distribution of final Pareto set are competitive compared with other evolutionary multi-objective optimizers.

关 键 词：群搜索算法 多目标优化 协同进化 上下文向量 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]