基于聚类的快速多目标遗传算法  

A Fast Multi-Objective Genetic Algorithm Based on Clustering

作　　者：郑金华[1] 史忠植[1] 谢勇[2] 

机构地区：[1]中国科学院计算技术研究所智能信息处理重点实验室,北京100080 [2]湘潭大学信息工程学院,湘潭411105

出　　处：《计算机研究与发展》2004年第7期1081-1087,共7页Journal of Computer Research and Development

基　　金：国家自然科学基金项目 ( 90 10 40 2 1) ;湖南省自然科学基金项目 ( 0 1JJY2 0 60 )

摘　　要：多目标遗传算法非常适合于求解多目标优化问题 讨论了进化个体之间的支配关系及有关性质 ,论证了可以用快速排序的方法对进化群体中的个体进行分类 ,同时探讨了用聚类方法来保持群体的多样性 ,具体讨论了基于层次凝聚距离的聚类 ,在此基础上提出了用分类和聚类的方法构造新的进化群体 理论分析与实验结果表明 。Multi-objective genetic algorithm based on Pareto optimum is much suitable for solving multi-objective optimization problems. In this paper,the relations between individuals and some features about these relations are discussed. It is proved that the individuals of an evolutionary population can be classified by the idea of quick sort. At the same time,the approach to maintain diversity of solutions by clustering algorithms is discussed,and the clustering algorithm based on hierarchical aggregation is also discussed. Then by using the quick sort algorithm and the clustering procedure,an algorithm of constructing a new evolutionary population is proposed. It is shown by theoretic analysis and experimental results that the convergent speed of the algorithm discussed is more efficient than the other existing algorithms.

关 键 词：聚类 多目标优化 多目标遗传算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]