首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于立队竞争的演化算法
引用本文:刘良兵,吴方才,黄樟灿.基于立队竞争的演化算法[J].武汉大学学报(理学版),2003,49(3):323-326.
作者姓名:刘良兵  吴方才  黄樟灿
作者单位:1. 华中师范大学,数学系,湖北,武汉,430070
2. 武汉理工大学,自动化学院,湖北,武汉,430070
3. 武汉大学,软件工程国家重点实验室,湖北,武汉,430072
基金项目:国家自然科学基金资助项目 ( 60 2 0 40 0 1)
摘    要:将竞争机制引入到变异算子中,从而使变异算子具有了很强的群体特性,同时为使算子更加简练和更具通用性,在这种竞争中作者更加注重个体适应值的相对次序而不是它们的绝对大小。基于这种思想提出了一种新的演化算法,在该算法中变异的作用已不仅仅是简单地恢复群体中损失的多样性,此时它已成为一种主要的搜索算子。通过实验发现新算法具有很高的搜索效率和求解精度。

关 键 词:立队竞争  演化算法  搜索算子  搜索效率  求解精度  演化计算  自适应性变异  竞争变异
文章编号:1671-8836(2003)03-0323-04
修稿时间:2002年12月16

Evolutionary Algorithms Based on Sort-Competition
LIU Liang-bing,WU Fang-cai,HUANG Zhang-can.Evolutionary Algorithms Based on Sort-Competition[J].JOurnal of Wuhan University:Natural Science Edition,2003,49(3):323-326.
Authors:LIU Liang-bing  WU Fang-cai  HUANG Zhang-can
Institution:LIU Liang-bing1,WU Fang-cai2,HUANG Zhang-can3
Abstract:In this paper, competition mechanism is drawn into mutation operator through which the mutation operator shows its high population property and the mutation operator itself becomes less complicated and general in use. In the competition proposed, more importance is attached to the relative order of the fitness than to their accurate values. Based on the above, a new evolutionary algorithm is presented in which the act of mutation is not the simple resumption of the lost population variety. Moreover, it becomes a main searching operator. The conclusion drawn through experiments is that the searching efficiency and solving accuracy of the new algorithm are quite high.
Keywords:competition  adaptive mutation  self-congregate
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号