Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (14): 60-62.DOI: 10.3778/j.issn.1002-8331.2009.14.017

• 研究、探讨 • Previous Articles     Next Articles

Hybrid genetic algorithm for TSP

YUAN Qi-zhao1,ZHU Yu-fei1,2,ZHENG Jin-hua1   

  1. 1.Institute of Information Engineering,Xiangtan University,Xiangtan,Hunan 411105,China
    2.College of Information Science and Engineering,Central South University,Changsha 410083,China
  • Received:2008-03-27 Revised:2008-06-03 Online:2009-05-11 Published:2009-05-11
  • Contact: YUAN Qi-zhao

求解TSP问题的混合遗传算法

袁琦钊1,朱云飞1,2,郑金华1   

  1. 1.湘潭大学 信息工程学院,湖南 湘潭 411105
    2.中南大学 信息科学与工程学院,长沙 410083
  • 通讯作者: 袁琦钊

Abstract: Traveling Salesman Problem(TSP) is a classical kind of nondeterministic polynomial completeness.Now these are many ways that can solve it,but a hybrid genetic algorithm is useful way for TSP.The normal hybrid genetic algorithms include genetic algorithm combined with steepest descent method and simulated annealing genetic algorithm.In the text,a greedy composite operator and climb operator are introduced for increasing the local searching ability.The experiment result demonstrates that the method is valid and effective.

Key words: Traveling Salesman Problem(TSP), genetic algorithm, steepest descent method, simulated annealing genetic algorithm, greedy composite operator, climb

摘要: TSP问题是一类经典的NP问题,目前有很多方法对其求解,而用混合遗传算法对其求解取得了很好的成效。常见的混合遗传算法有遗传算法与最速下降法相结合(GACSDM)、遗传算法与模拟退火法相结合(SAGA)。设计了贪婪的复合变异算子(GCM),并引入隔代爬山法算子(Climb)增加遗传算法的局部搜索能力。实验结果表明该算法是有效的。

关键词: 旅行商问题, 遗传算法, 最速下降法, 模拟退火法, 贪婪的复合变异算子, 爬山法