基于粒子群和声搜索混合算法的多模态函数优化  

Multi-modal Function Optimization Based on Particle Swarm Harmony Search Hybrid Algorithm

作　　者：黄金山[1] 王东风[1] 

机构地区：[1]华北电力大学控制与计算机工程学院,河北保定071003

出　　处：《电力科学与工程》2013年第7期33-38,共6页Electric Power Science and Engineering

基　　金：中央高校基本科研业务费专项资金资助项目(11MG49)

摘　　要：和声搜索算法是一种启发式全局搜索算法,概念简单、参数少、易于实施,加强了鲁棒性和基础搜索机制的灵活性。针对粒子群算法和和声搜索算法各自的特点提出了一种新的搜索算法———粒子群和声搜索混合算法(PSO-IHS)。新算法将和声搜索算法的和声库初始解通过粒子群(PSO)算法进行改进产生,同时对和声搜索算法的算法参数和每次迭代产生的新解个数也作了改进,并分别用5个多模态测试函数进行了仿真,用于验证算法的搜索性能。仿真结果表明,粒子群和声搜索混合算法提高了函数优化的搜索效率,具有较高的寻优性能和较强的跳出局部极小的能力。Harmony Search（HS） algorithm is a kind of heuristic global search algorithm that was developed few years ago.HS has many advantages such as simple concept,few parameters to be tuned and easy to implement.It is robust and flexible in searching mechanism.Based on the characteristics of Particle Swarm Optimization（PSO） and improved HS,this paper presentes a hybrid algorithm that is called PSO-IHS.PSO-IHS improves the initial solution of harmony memory of HS by using PSO.PSO-IHS improves the generation method of new solutions of HS.Optimization test of five multimodal functions show that the proposed new algorithm is feasible and effective.It not only improves the searching efficiency but also has stronger ability to escape from local minimum.

关 键 词：和声搜索算法 粒子群算法 多模态函数 智能优化算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]