一种基于相似个体的多目标进化算法  

Multi-objective evolutionary algorithm based on similar individuals

作　　者：伍军[1] 郑金华[1] 文诗华[1] 

机构地区：[1]湘潭大学信息工程学院,湖南湘潭411105

出　　处：《计算机工程与应用》2008年第14期67-72,84,共7页Computer Engineering and Applications

基　　金：国家自然科学基金(the National Natural Science Foundation of China under Grant No.60773047);国家高技术研究发展计划(863)(the Na-tional High-Tech Research and Development Plan of China under Grant No.2001AA114060);教育部留学回国人员科研启动基金(The Pro-ject-sponsored by SRF for ROCS;SEMNo.教外司留[2005]546);湖南省自然科学基金(the Natural Science Foundation of Hunan Province of China under Grant No.05JJ30125);湖南省教育厅重点科研项目(No.06A074)

摘　　要：分布性保持是多目标进化算法研究的一个重要方面,一个好的分布性能给决策者提供更多合理有效的选择。Pareto最优解的分布性主要体现在分布广度与均匀性两个方面。提出一种基于相似个体的多目标进化算法(SMOEA)。在种群维护中删除相似程度最大的个体;在进化操作中,选取了相似程度最大的个体进行进化。与目前经典算法NSGA-II和ε-MOEA进行比较,结果表明新算法拥有良好的分布性,同时也较好的改善了收敛性。Diversity maintenance strategy is an important part of studies on multi-objective evolutionary Mgorithms （ MOEAs ）. A good diversity can give a decision-maker more reasonable and efficient selections. The diversity of pareto optimal solutions include the span and the uniformity of solutions. This paper proposes a multi-objective evolutionary algorithm based on similar individuals （ SMOEA ）. In the process of population maintenance, it deletes individuals with the most similarity. While in the process of evolution operations, the most similar individuals are selected to evolve. Compared with NSGA-II and ε-MOEA, the experimental results demonstrate that the new algorithm can converge to the true Pareto front fast and can obtain good convergence at the same time.

关 键 词：多目标进化算法 多目标优化问题 种群维护 分布性 相似个体 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]