一种基于多亲遗传机制的多目标优化算法  

A MULTI-OBJECTIVE GENETIC ALGORITHM BASED ON MULTI-PARENT CROSSOVER

作　　者：吴佳英[1] 李平[1] 郑金华[2] 

机构地区：[1]长沙理工大学计算机与通信工程学院,湖南长沙410076 [2]湘潭大学信息工程学院,湖南湘潭411105

出　　处：《计算机应用与软件》2008年第2期52-53,79,共3页Computer Applications and Software

基　　金：湖南省自科基金项目(05JJ30125);湖南省教育厅基金项目(06B005)

摘　　要：给出了进化个体之间的关系和非支配集中不同个体之间的相关性质,参考快速排序的思想,提出了一种有效的构造非支配集的算法。在此基础上,将多亲遗传算法与改进的快速排序构造非支配集的算法相结合,提出了一种基于多亲遗传机制的多目标优化算法。最后对提出算法进行了分析,采用了测试函数进行了仿真实验,获得了理想的实验结果。The relation between individuals and some features of these relations are discussed. By referring to the idea of quick sort, a valid algorithm of non-dominated set construction is proposed. Then with the combination of the multi-parent genetic algorithm and the improved non -dominated set construction algorithm based on quick sort, a Multi-objective Genetic Algorithm based on Multi-parent Crossover（MOGAMC） is put forward. Finally, analysis is made, and simulative experimental figures show that the algorithm has nice performance.

关 键 词：多亲遗传算法 多目标优化 非支配集 

分 类 号：O242.23[理学—计算数学] Q943[理学—数学]