• 论文 •    

一种改进的周期性病毒进化遗传算法应用研究

梁旭,王佳,黄明   

  1. 大连交通大学 软件学院,辽宁大连116028
  • 出版日期:2007-04-15 发布日期:2007-04-25

Application research on improved cyclical virus evolution genetic algorithm

LIANG Xu, WANG Jia, HUANG Ming   

  1. Sch. of Software Tech., Dalian Jiaotong Univ., Dalian116028, China
  • Online:2007-04-15 Published:2007-04-25

摘要: 为不断更新传统病毒进行遗传算法的进化病毒群体,提出了一种改进的周期性病毒进化遗传算法。横向上通过周期性生成病毒、周期性进行病毒感染操作,有效地增加了种群的多样性,并减少了计算工作量,提高搜索效率。纵向上对主群体采用逆二分法选择、循环按趟交叉的机制,增强了算法的收敛性能,加快了收敛速度。将提出的改进算法用Muth and Thompson基准问题测试,显示了该算法的有效性。

关键词: 周期性病毒, 逆二分法选择, 循环按趟交叉, 遗传算法

Abstract: Aiming at the insufficiency of traditional Virus Evolution Genetic Algorithm (VEGA) both in unceasing renewing and evolving virus group and the heavy workload, an improved Cyclical Virus Evolution Genetic Algorithm (CVEGA) was proposed. Crosswise, it could effectively enhance population-diversification, reduce workload and improve the efficiency by cyclical producing virus and carrying on the virus. Longitudinal, it shrank the searching space and increased convergence performance and speed as well, via using counter dichotomy selection and circulation cross according to time to the host community. This improved GA has been tested by benchmark problem in Muth and Thompson, and its validity was revealed.

Key words: cyclical virus, counter dichotomy selection, circulation cross according to time, genetic algorithm

中图分类号: