摘要
随着工程技术的发展与优化问题数学模型的完善,许多优化问题从低维优化发展成高维的大规模复杂优化,成为实值优化领域的一个热点问题.通过对大规模问题的特点分析,提出了随机动态的协同进化策略,将其加入动态多种群粒子群优化算法中,实现了对种群粒子和决策变量的双重分组.最后,使用CEC2013的大规模全局优化算法的测试集对新算法进行测试,通过和其他算法的对比,验证算法的有效性.
With the development of engineering technology and the improvement of mathematical model,a large number of optimization problems have been developed from low dimensional optimization to large-scale complex optimization. Large scale global optimization is an active research topic in the real-parameter optimization. Based on the analysis of the characteristics of large scale problems,a stochastic dynamic cooperative coevolution strategy is proposed in the article. Additionally,a strategy is added to the dynamic multi-swarm particle swarm optimization algorithm. Then,the dual grouping of population and decision variables is realized. Next,the performance of the novel optimization on the set of benchmark functions provided for the CEC2013 special session on large scale optimization is reported. Finally the validity of the algorithm is verified by comparing with other algorithms.
作者
梁静
刘睿
于坤杰
瞿博阳
LIANG Jing;LIU Rui;YU Kun-Jie;QU Bo-Yang(School of Electrical Engineering,Zhengzhou University,Zhengzhou 450001,China;School of Electric and Information Engineering,Zhongyuan University of Technology,Zhengzhou 450007,China)
出处
《软件学报》
EI
CSCD
北大核心
2018年第9期2595-2605,共11页
Journal of Software
基金
国家自然科学基金(61673404
61473266
61876169)
中国博士后科学基金(2017M622373)~~
关键词
大规模全局优化算法
动态多种群粒子群优化算法
协同进化
基准测试函数
large scale global optimization
dynamic multi-swarm particle swarm optimization
cooperative coevolution
benchmark function