作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2010, Vol. 36 ›› Issue (14): 166-168. doi: 10.3969/j.issn.1000-3428.2010.14.060

• 人工智能及识别技术 • 上一篇    下一篇

基于改进蚁群算法的纳什均衡求解

王志勇1,韩 旭1,许维胜1,杨继君2   

  1. (1. 同济大学电子与信息工程学院,上海 201804;2. 同济大学经济与管理学院,上海 201804)
  • 出版日期:2010-07-20 发布日期:2010-07-20
  • 作者简介:王志勇(1984-),男,硕士研究生,主研方向:优化算法,智能控制;韩 旭,博士;许维胜,教授、博士、博士生导师;杨继君,博士
  • 基金资助:
    国家自然科学基金资助项目(70871091)

Nash Equilibrium Solution Based on Improved Ant Colony Algorithm

WANG Zhi-yong1, HAN Xu1, XU Wei-sheng1, YANG Ji-jun2   

  1. (1. School of Electronics and Information Engineering, Tongji University, Shanghai 201804;2. School of Economic & Management, Tongji University, Shanghai 201804)
  • Online:2010-07-20 Published:2010-07-20

摘要: 在基本蚁群算法寻优机制的基础上,提出一种用于求解有限n人非合作博弈的纳什均衡解的改进蚁群算法。在全局搜索中,引入遗传算法中的交叉和变异操作提高算法的全局搜索能力。在局部搜索中,嵌入动态随机搜索技术使算法加速收敛到最优解,并通过引入控制步长调整随机搜索向量,保证蚁群始终在混合策略空间内。算例测试结果表明,与传统的遗传算法相比,该算法具有更好的计算性能。

关键词: 蚁群算法, 非合作博弈, 纳什均衡

Abstract: This paper presents an improved ant colony algorithm for solving Nash equilibrium of n persons’ non-cooperative games, using the basic principle of ACO algorithm. In the global search phase, crossover operation and mutation operation of GA are incorporated to improve global exploring ability. During the local search phase, dynamic random search technique is embedded to accelerate the convergence to the optimal solution, and the random search vector is adjusted by controlling step length to ensure search in the mixed strategy profile space. Numerical experiments show that the proposed algorithm has better performance than the traditional genetic algorithm.

Key words: ant colony algorithm, non-cooperative game, Nash equilibrium

中图分类号: