一种军棋机器博弈的多棋子协同博弈方法  

A multi-chess collaborative game method for military chess game machine

作　　者：张小川 王宛宛 彭丽蓉[3] ZHANG Xiaochuan;WANG Wanwan;PENG Lirong(Liangjiang Institute of Artificial Intelligence,Chongqing University of Technology,Chongqing 400054,China;School of Computer Science and Engineering,Chongqing University of Technology,Chongqing 400054,China;Faculty Information Engineering,Chongqing Industry Polytechnic College,Chongqing 401120,China)

机构地区：[1]重庆理工大学两江人工智能学院,重庆400054 [2]重庆理工大学计算机科学与工程学院,重庆400054 [3]重庆工业职业技术学院信息工程学院,重庆401120

出　　处：《智能系统学报》2020年第2期399-404,共6页CAAI Transactions on Intelligent Systems

基　　金：国家自然科学基金项目(61702063);重庆理工大学研究生创新基金项目(ycx2018244)。

摘　　要：针对在军棋博弈不完全信息对弈中,面对棋子不同价值、不同位置、不同搭配所产生的不同棋力,传统的单子意图搜索算法,不能满足棋子之间的协同性与沟通性,同时也缺乏对敌方的引诱与欺骗等高级对抗能力。本文提出一种结合UCT搜索策略的高价值棋子博弈方法,实现高价值棋子协同博弈的策略。实战经验表明:高价值多棋子军棋协同博弈策略优于单棋子军棋博弈策略。Owing to incomplete information on the military chess and the different strengths of chess pieces with different values,positions,and combinations,the traditional single-intention search algorithm cannot satisfy the coordination and communication requirements of chess pieces and lacks advanced confrontation capabilities,such as temptation and deception of the enemy.This study proposes the combination of the high-value chess piece game method and the UCT search strategy to achieve a high-value chess piece cooperative game strategy that can be used to solve the problems of the military chess game.Practical experience shows that the high-value multipiece military chess game strategy is better than the high-value single-piece military chess game strategy.

关 键 词：机器博弈 军棋 协同博弈 Q学习算法 攻守平衡 维度灾难 UCT 高价值棋子 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]