计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (29): 72-74.DOI: 10.3778/j.issn.1002-8331.2009.29.020

• 研究、探讨 • 上一篇    下一篇

非劣解分布范围的度量——S-度量

马光娟1,王宇平2   

  1. 1.西安电子科技大学 理学院,西安 710071
    2.西安电子科技大学 计算机学院,西安 710071
  • 收稿日期:2008-05-22 修回日期:2008-09-11 出版日期:2009-10-11 发布日期:2009-10-11
  • 通讯作者: 马光娟

S-Measure:Extensive measure of non-dominated solutions for multiobjective programming

MA Guang-juan1,WANG Yu-ping2   

  1. 1.Department of Science,Xidian University,Xi’an 710071,China
    2.Department of Computer Science,Xidian University,Xi’an 710071,China
  • Received:2008-05-22 Revised:2008-09-11 Online:2009-10-11 Published:2009-10-11
  • Contact: MA Guang-juan

摘要: 提出了一种新的非劣解前端宽广性的度量,S-度量。(1)粗略估计问题I的边界解的集合;(2)由二水平正交设计的思想,从这个集合中选取指定分布比较均匀的参考解;(3)从非劣解集中找与每个参考解最近的解,并计算其距离;(4)给出S-度量的定义.将S-度量与其他一些非劣解质量的度量相结合,从而可以对多目标遗传算法从多个角度进行评价和比较.

关键词: 多目标优化, 非劣解, 正交设计, S-度量

Abstract: This paper proposes an extensive measure for the non-dominated solutions,S-Measure.(1)Find the border solution set for problem I;(2)from the two levels orthoplan,select some reference solutions from the border solution set;(3)find the nearest solutions for each reference solution from the non-dominated solutions,and calculate its distance;(4)give the definition of S-Measure.S-Measure can be applied to complement the other quality measures in order to evaluate and compare multiobjective programming algorithms from different perspective.

Key words: multiobjective programming, non-dominated solutions, orthogonal design, S-Measure

中图分类号: