摘要
群搜索优化算法(Group Search Optimizer,GSO)是一类基于发现者-加入者(Producer-Scrounger,PS)模型的新型群体随机搜索算法。尽管该算法在解决众多问题中表现优越,但其依然面临着早熟和易陷入局部最优的问题,为此,提出了一种基于一般反向学习策略的群搜索优化算法(GOGSO)。该算法利用反向学习策略来产生反向种群,然后对当前种群和反向种群进行精英选择。通过对比实验表明,该方法效果良好。
Group search optimizer（GSO）is a new swarm intelligence algorithms based on the producer-scrounger model.GSO has been shown to yield good performance for solving various optimization problems.However,it tends to suffer from premature convergence and get stuck in local minima.This paper proposed an enhanced GSO algorithm called GOGSO,which employs generalized opposition-based learning to transform the current population into a new opposition population and uses an elite selection mechanism on the two populations.Experiments were conducted on a comprehensive set of benchmark functions.The results show that OGSO obtains promising performance.
出处
《计算机科学》
CSCD
北大核心
2012年第9期183-187,共5页
Computer Science
基金
国家自然科学基金(60975050
61070243
61165004)
高等学校博士学科点专项科研基金(20070486081)
中央高校基本科研业务费专项资金(6081014)
河北省科技支撑计划项目(11213587)
江西省自然科学基金(20114BAB201025)
江西省教育厅科技项目(GJJ12307)资助
关键词
群搜索优化算法
反向学习
数值优化
Group search optimizer
Opposition-based learning
Numerical optimization
