计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (15): 36-38.

• 理论研究 • 上一篇    下一篇

一种基于个体密度估算的多目标优化演化算法

敖友云1,迟洪钦2   

  1. 1.安庆师范学院 计算机与信息学院,安徽 安庆 246001
    2.上海师范大学 数理信息学院,上海 200234
  • 收稿日期:2007-09-04 修回日期:2007-11-19 出版日期:2008-05-21 发布日期:2008-05-21
  • 通讯作者: 敖友云

Multi-objective optimization evolutionary algorithm based on individual density estimation

AO You-yun1,CHI Hong-qin2   

  1. 1.School of Computer and Information,Anqing Teachers College,Anqing,Anhui 246001,China
    2.Mathematics and Science College,Shanghai Normal University,Shanghai 200234,China
  • Received:2007-09-04 Revised:2007-11-19 Online:2008-05-21 Published:2008-05-21
  • Contact: AO You-yun

摘要: 通过在目标空间中利用目标本身信息估算个体k最近邻距离之和,作为个体的密度信息,根据个体的密度信息对群体中过剩的非劣解进行逐个去除,以便更好地维护解的多样性,由此给出了一种基于个体密度估算的多目标优化演化算法IDEMOEA。用这个算法对几个典型的多目标优化函数进行测试。测试结果表明,算法IDEMOEA求解多目标优化问题是行之有效的。

关键词: 演化算法, 多目标优化, 多样性维护, Pareto最优

Abstract: A multi-objective optimization evolutionary algorithm based on Individual Density Estimation(IDEMOEA) is presented,which use the k nearest neighbors for individual density estimation by the information of objectives themselves in such a way that distances to the k nearest neighbors are summed together,namely individual density information,while prune the over-plus of non-dominated solutions in the population one by one according to individual density information with the purpose of preserving the diversity of solutions.A few of benchmark multi-objective optimization functions are tested.Experimental results demonstrate that the algorithm developed is efficient for solving multi-objective optimization problems.

Key words: evolutionary algorithm, multi-objective optimization, diversity maintenance, Pareto optimal