Computer Engineering and Applications ›› 2021, Vol. 57 ›› Issue (24): 198-204.DOI: 10.3778/j.issn.1002-8331.2007-0322

• Pattern Recognition and Artificial Intelligence • Previous Articles     Next Articles

Ant Colony Optimization for Continuous Domains Applied to Cooperative Game

LI Zhuangkuo, CHANG Kaixuan   

  1. School of Business, Guilin University of Electronic Technology, Guilin, Guangxi 541004, China
  • Online:2021-12-15 Published:2021-12-13

合作博弈的连续蚁群算法求解

李壮阔,常凯旋   

  1. 桂林电子科技大学 商学院,广西 桂林 541004

Abstract:

In the theoretical research of cooperative games, the classic solution concept of cooperative game does not embody the limited rationality and interactional game-playing behavior of the player in the solution process. In the real world, the allocation schemes are often formed through the player’s rational interaction and strategic changes. This paper introduces rational factors and control factors to describe the players’ decision-making behaviors during the game, establishes a cooperative game model considering interaction behavior, and uses ant colony optimization for continuous domains to solve the cooperative game. An example shows that this method can ensure the allocation scheme satisfies the effectiveness and the individual rationality, and can quickly obtain the only allocation scheme of the alliance. It provides a new idea and tool for solving cooperative games.

Key words: cooperative game, interactive behavior, negotiation, ant colony optimization for continuous domains

摘要:

在合作博弈的理论研究中,经典的合作博弈解概念在求解过程中没有体现出局中人的有限理性和互动博弈行为。而在现实博弈环境中,联盟的分配方案更多是通过局中人间理性互动与策略博弈形成的。引入理性因子和控制因子来描述局中人在博弈过程中的决策行为,建立了考虑互动行为的合作博弈模型,并利用连续蚁群算法对合作博弈进行求解。算例表明该解法可以保证分配方案满足有效性和个体理性,并能快速得到联盟的唯一分配方案。这为合作博弈的求解提供了新的思路与工具。

关键词: 合作博弈, 互动行为, 谈判, 连续蚁群算法