一种改进的和声搜索算法及其应用  

Improved Harmony Search Algorithm and Its Application

作　　者：张敬敏[1] 秦彭[1] 贺毅朝[1] 王彦祺[1] 

机构地区：[1]石家庄经济学院信息工程学院,石家庄050031

出　　处：《科学技术与工程》2013年第13期3763-3767,3773,共6页Science Technology and Engineering

基　　金：河北省科技攻关项目(11213525D);石家庄经济学院预研05资助

摘　　要：为了克服和声搜索算法(Harmony Search Algorithm,HSA)方向性差和收敛速度慢的缺陷,提出一种新的和声搜索算法。新算法设置一个主种群和多个子种群,主种群和子种群分别采用全局差分HSA和局部差分HSA进行优化,所有种群通过排序循环划分方法进行协同优化。随机3—SAT测试实例和三个典型Benchmark函数上的实验表明,新算法是可行的、有效的。In order to overcome harmony search algorithm defect of the directional difference and slow conver- gence speed, a new harmony search algorithm is put forward. The new algorithm sets a main population and more than one child population, the main population and child population respectively using global difference HSA and local difference HSA optimization, all population collaborative optimization was through the cycle division method.The test case of random 3--SAT and three classical Benchmark functions show that the new algorithm is feasible and effective.

关 键 词：和声搜索算法 差分演化算法 协同 循环划分 可满足性问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]