自适应改进和声—单纯形进化算法研究

摘要

For the purpose of avoiding the disadvantage of harmony search algorithm, this paper proposed an adaptive improved harmony pattern search (AIHSEA) algorithm. In AIHSEA, it added a mutation strategy to enhance the robustness of HS. In order to increase the directivity of the harmony search, it implemented the simplex method timely. It adjusted its global and local searches by using adaptive parameters HMCR, PAR and BW. Finally, AIHSEA was tested by six standard benchmark functions and contrasted with HS, HIS and GHS algorithm. The test results show the favorable abilities of accuracy and escaping local minimums.%针对和声搜索算法的不足,提出了一种自适应改进和声-单纯形进化算法(AIHSEA).通过在新算法中加入变异策略对和声微调进行改进来增强算法的鲁棒性；适时执行单纯形算子增加群体搜索的方向性来加快搜索；采用自适应参数HMCR、PAR和BW调节全局和局部搜索.采用六个标准的优化算法测试函数对AIHSEA进行测试,并与HS、IHS和GHS算法进行对比,仿真结果表明AIHSEA算法具有较强的精确寻优和跳出局部最优的能力.