多核CPU下基因库的多种群遗传算法  

A gene pool multi-population genetic algorithm based on multi-core CPU

作　　者：孙如祥[1,2] 黄柏雄[1] 谢祥宏[1] 夏曼[1,2] 

机构地区：[1]广西大学计算机与电子信息学院,广西南宁530004 [2]广西职业技术学院计算机与电子信息工程系,广西南宁530226

出　　处：《传感器与微系统》2011年第8期135-137,共3页Transducer and Microsystem Technologies

基　　金：国家自然科学基金资助项目(50605010)

摘　　要：近年来,多种群遗传算法被广泛地应用到各领域中,鉴于该算法的有效性,在其基础上提出了一种用于在多核CPU环境下求解TSP问题的多种群遗传算法,利用接收到的最优个体构建基因库,种群之间通过共享基因库来实现种群间信息的交流。通过对TSP问题求解的数值实验表明:提出的算法提高了求解质量,而且在效率上也具有很大的优势。In recent years,multi-population genetic algorithm has been widely applied in various fields.In view of validity of the algorithm,a multi-population genetic algorithm for TSP is proposed based on multi-core CPU environment,the gene pool is constructed by using the best individuals of every populations,the information exchange between populations is achieved by the shared gene pool.The experimental results of TSP show that the proposed algorithm improves the solution quality and also has a great advantages in efficiency.

关 键 词：多核 基因库 多种群 遗传算法 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]