兵棋推演的智能决策技术与挑战  

Intelligent Decision Making Technology and Challenge of Wargame

作　　者：尹奇跃 赵美静[1] 倪晚成[1,2] 张俊格 黄凯奇[1,2] YIN Qi-Yue;ZHAO Mei-Jing;NI Wan-Cheng;ZHANG Jun-Ge;HUANG Kai-Qi(Institute of Automation,Chinese Academy of Sciences,Beijing 100190;University of Chinese Academy of Sciences,Beijing 100049)

机构地区：[1]中国科学院自动化研究所,北京100190 [2]中国科学院大学,北京100049

出　　处：《自动化学报》2023年第5期913-928,共16页Acta Automatica Sinica

基　　金：国家自然科学青年基金(61906197)资助。

摘　　要：近年来,以人机对抗为途径的智能决策技术取得了飞速发展,人工智能(Artificial intelligence, AI)技术AlphaGo、AlphaStar等分别在围棋、星际争霸等游戏环境中战胜了顶尖人类选手.兵棋推演作为一种人机对抗策略验证环境,由于其非对称环境决策、更接近真实环境的随机性与高风险决策等特点,受到智能决策技术研究者的广泛关注.通过梳理兵棋推演与目前主流人机对抗环境(如围棋、德州扑克、星际争霸等)的区别,阐述了兵棋推演智能决策技术的发展现状,分析了当前主流技术的局限与瓶颈,对兵棋推演中的智能决策技术研究进行了思考,期望能对兵棋推演相关问题中的智能决策技术研究带来启发.In recent years,decision-making intelligence based on human-machine confrontation has achieved rapid development.For example,artificial intelligence(AI)technology such as AlphaGo and AlphaStar have defeated top human players in games Go and StarCraft,respectively.Nowadays,wargame,as a new verification environment for human-machine confrontation,attracts more and more researchers due to new challenges being raised,i.e.,asymmetric environmental decision-making and randomness with high-risk decision-making.In this paper,we will sort out the differences between wargame and the current mainstream human-machine confrontation environments such as Go,Poker and StarCraft.Then,we explain the development status of wargame intelligent technology,and analyze the limitations of current mainstream technologies.Finally,we present our thoughts about future development of technologies for wargame,hoping to inspire researchers for through study on wargame.

关 键 词：兵棋推演 人机对抗 智能决策技术 博弈学习 

分 类 号：O225[理学—运筹学与控制论] TP18[理学—数学]