一种求解多模态复杂问题的混合和声差分算法

摘要

This paper presents a hybrid algorithm (HHSDE) based on harmony search and differential evolution for solv-ing multi-modal complex optimization. In different evolution stages, HHSDE algorithm self-adaptively selects harmony search (HS) or differential evolution (DE) algorithm as the means of updating the next generation of population on basis of the cumulative success rate of weighted update, in addition, it changes the mutation strategy of differential evolution (DE) algorithm for balancing the global and local search ability of the differential evolution (DE) algorithm. To investig-ate the performance of HHSDE, ten multi-modal Benchmark functions were tested. The experimental results, compared with other algorithms by Wilcoxon rank sum test, indicate that HHSDE algorithm has the advantages such as fast con-vergence speed, high solution precision and excellent stability.%针对多模态复杂优化问题,提出了一种基于和声搜索和差分进化的混合优化算法:HHSDE算法.在不同的进化阶段,HHSDE算法依据累积加权更新成功率来自适应地选择和声算法或差分算法作为更新下一代种群的方式,并改进了差分算法的变异策略来平衡差分算法的全局与局部搜索能力.通过对10个多模态Benchmark函数进行测试,利用Wilcoxon秩和检验对不同算法的计算结果进行比较,结果表明HHSDE算法具有收敛速度快,求解精度高,稳定性好等优势.