群队级兵棋实体智能行为决策方法研究

doi:10.12305/j.issn.1001-506X.2022.08.21

系统工程与电子技术 ›› 2022, Vol. 44 ›› Issue (8): 2562-2569.doi: 10.12305/j.issn.1001-506X.2022.08.21

群队级兵棋实体智能行为决策方法研究

刘满^1,², 张宏军^1,*, 徐有为¹, 冯欣亮³, 冯玉芳⁴

1. 陆军工程大学指挥控制工程学院, 江苏南京 210007
2. 中国人民解放军73131部队, 福建漳州 363000
3. 陆军步兵学院, 江西南昌 330000
4. 中国人民解放军71375部队, 黑龙江哈尔滨 150000

收稿日期:2021-05-26 出版日期:2022-08-01 发布日期:2022-08-24
通讯作者: 张宏军
作者简介:刘满(1986—), 男, 博士研究生, 主要研究方向为深度学习、强化学习、军事智能决策和系统仿真|张宏军(1963—), 男, 教授, 博士研究生导师, 主要研究方向为数据与知识工程、计算机仿真理论|徐有为(1994—), 男, 博士研究生, 主要研究方向为知识图谱、图神经网络|冯欣亮(1986—), 男, 讲师, 硕士, 主要研究方向为军事指挥和军事体育|冯玉芳(1981—), 女, 工程师, 博士, 主要研究方向为信息处理与模式识别
基金资助:
国家自然科学基金(61806221)

Research on behavior decision-making of multi entities in group-level wargame

Man LIU^1,², Hongjun ZHANG^1,*, Youwei XU¹, Xinliang FENG³, Yufang FENG⁴

1. Command and Control Engeering College, Army Engineering University, Nanjing 210007, China
2. Unit 73131 of the PLA, Zhangzhou 363000, China
3. Army Infantry Academy of the PLA, Nanchang 330000, China
4. Unit 71375 of the PLA, Harbin 150000, China

Received:2021-05-26 Online:2022-08-01 Published:2022-08-24
Contact: Hongjun ZHANG

摘要/Abstract

摘要：

计算机兵棋能够对敌对双方或多方的军事行动进行模拟, 为军事智能决策的研究提供贴近真实战争的决策背景和试验环境。随着被控制棋子类型和数量增加, 这些作战实体的行为决策控制将面临巨大的挑战。本文提出了知识与数据互补的兵棋行为决策算法, 针对较多作战实体(50个左右)的群队级兵棋, 提出了按实体类型分组决策、作战目标指引、位置评估指标并行计算等方法, 设计了群队级兵棋人工智能(artificial intelligence, AI)技术框架, 并实现了一个群队级兵棋AI。该兵棋AI将知识、数据和学习融为一体, 具有设计简单、迁移能力强、决策灵活等优点。

关键词: 群队级兵棋, 并行计算, 行为决策, 多实体控制

Abstract:

Computer wargame can simulate the military actions of two or more opposed parties, which provides a decision-making background and experimental environment close to the real war for the research of military intelligent decision-making. With the increase of the type and number of controlled entities, the intelligent behavior decision-making will face great challenges. This paper proposes the behavior decision-making algorithm by complementary knowledge and data for group-level wargame which has multi entities (about 50), including decision-making grouped as entity type, operational target allocation guidance, parallel calculation of position evaluation. Then, we design the framework of group-level wargame artificial intelligence (AI) technology, and realizes it. The wargame AI integrates knowledge, data and learning, has the advantages of less design workload, strong migration ability and flexible decision-making.

Key words: group-level wargame, parallel computing, behavior decision-making, multi entities control

中图分类号:

TP273

刘满, 张宏军, 徐有为, 冯欣亮, 冯玉芳. 群队级兵棋实体智能行为决策方法研究[J]. 系统工程与电子技术, 2022, 44(8): 2562-2569.

Man LIU, Hongjun ZHANG, Youwei XU, Xinliang FENG, Yufang FENG. Research on behavior decision-making of multi entities in group-level wargame[J]. Systems Engineering and Electronics, 2022, 44(8): 2562-2569.

图/表 7

表1

图1

图2

图3

图4

图5

表2

参考文献 25

1	YE D H, LIU Z F, SUN M, et al. Mastering complex control in moba games with deep reinforcement learning[C]//Proc. of the AAAI Conference on Artificial Intelligence, 2020, 34(4): 6672- 6679.
2	VINYALS O , BABUSCHKIN I , GZAMECKI W M , et al. Grandmaster level in StarCraft Ⅱ using multi-agent reinforcement learning[J]. Nature, 2019, 575 (7782): 1- 5.
3	胡晓峰, 齐大伟. 智能决策问题探讨——从游戏博弈到作战指挥, 距离还有多远[J]. 指挥与控制学报, 2020, 6 (4): 356- 363.
	HU X F , QI D W . On problems of intelligent decision-making—how far is it from game-playing to operational command[J]. Journal of Command and Control, 2020, 6 (4): 356- 363.
4	张振, 黄炎焱, 张永亮, 等. 基于近端策略优化的作战实体博弈对抗算法[J]. 南京理工大学学报, 2021, 45 (1): 77- 83.
	ZHANG Z , HUANG Y Y , ZHANG Y L , et al. Battle entity confrontation algorithm based on proximal policy optimization[J]. Journal of Nanjing University of Science and Technology, 2021, 45 (1): 77- 83.
5	李琛, 黄炎焱, 张永亮, 等. Actor-Critic框架下的多智能体决策方法及其在兵棋上的应用[J]. 系统工程与电子技术, 2021, 43 (3): 755- 762.
	LI C , HUANG Y Y , ZHANG Y L , et al. Multi-agent decision-making method based on Actor-Critic framework and its application in wargame[J]. Systems Engineering and Electronics, 2021, 43 (3): 755- 762.
6	JONES R M , LAIRD J E , NIELSEN P E , et al. Automated intelligent pilots for combat flight simulation[J]. AI Magazine, 1999, 20 (1): 27- 42.
7	GRATCH J. Task-decomposition planning for command decision making[C]//Proc. of the 6th Conference on Computer Generated Forces and Behavioral Representation, 1996.
8	郑书奎, 吴琳, 贺筱媛. 基于深度学习的兵棋演习数据特征提取方法研究[J]. 指挥与控制学报, 2016, 2 (3): 194- 201.
	ZHENG S K , WU L , HE X Y . The feature extraction method of wargame exercises data based on deep learning[J]. Journal of Command and Control, 2016, 2 (3): 194- 201.
9	张可, 郝文宁, 余晓晗, 等. 基于遗传模糊系统的兵棋推演关键点推理方法[J]. 系统工程与电子技术, 2020, 42 (10): 2303- 2311. doi: 10.3969/j.issn.1001-506X.2020.10.19
	ZHANG K , HAO W N , YU X H , et al. Research on wargame key point reasoning method based on genetic fuzzy system[J]. Systems Engineering and Electronics, 2020, 42 (10): 2303- 2311. doi: 10.3969/j.issn.1001-506X.2020.10.19
10	PAN Y, NI W C, YANG Y. An algorithm to estimate enemy's location in WarGame based on pheromone[C]//Proc. of the 33rd Youth Academic Annual Conference of Chinese Association of Automation, 2018: 749-753.
11	田忠良, 刘昊. 智能算法在兵棋对抗推演中的应用[J]. 指挥控制与仿真, 2021, 43 (1): 40- 47.
	TIAN Z L , LIU H . Application of intelligent algorithm in military chess antagonism deduction[J]. Command Control & Simulation, 2021, 43 (1): 40- 47.
12	刘满, 张宏军, 郝文宁, 等. 战术级兵棋实体作战行动智能决策算法[J]. 控制与决策, 2020, 35 (12): 2977- 2985.
	LIU M , ZHANG H J , HAO W N , et al. Intelligent decision-making method of tactical-level wargames[J]. Control and Decision, 2020, 35 (12): 2977- 2985.
13	JAMES F D . The complete wargames handbook[M]. New York: William Morrow&Company, 1997: 1- 2.
14	U.S. The joint theater level simulation (JTLS) analyst's guide[R]. USA: 116 Lake View Parkway Suffolk, VA, 2007: 2435-2697.
15	CERANOWICZ A. Modular semi-automated forces[C]//Proc. of the 26th Conference on Winter Simulation, 1994: 755- 761.
16	BIMSON K, MARSDEN C, MCKENZIE F, et al. Knowledge-based tactical decision making in the CCTT SAF Prototype[C]//Proc. of the 4th Conference on Computer Generated Forcesand Behavioral Representation, 1994.
17	GIAMPAPA J A , SYCARA K P , OWENS S , et al. Extending the one SAF testbed into a C4ISR testbed[J]. Simulation, 2004, 80 (12): 681- 691. doi: 10.1177/0037549704050348
18	LIM C U, BAUMGARTEN R, COLTON S. Evolving behaviour trees for the commercial game DEFCON[C]//Proc. of the European Conference on the Applications of Evolutionary Computation, 2010: 100-110.
19	YOONES A S . Behavior trees for computer games[J]. International Journal of Artificial Intelligence Tools: Architectures, Languages, Algorithms, 2017, 26 (2): 1730001. doi: 10.1142/S0218213017300010
20	石俊杰. 基于有限状态机的游戏角色控制系统设计与实现[D]. 武汉: 华中科技大学, 2016.
	SHI J J. Design and implementation of a FSM-based role control system[D]. Wuhan: Huazhong University of Science & Technology, 2016.
21	蒋宗礼, 姜守旭. 形式语言与自动机理论[M]. 北京: 清华大学出版社, 2003.
	JIANG Z L , JIANG S X . Formal language and automata theory[M]. Beijing: Tsinghua University Press, 2003.
22	刘瑞峰, 王家胜, 张灏龙, 等. 行为树技术的研究进展与应用[J]. 计算机与现代化, 2020, (2): 76- 82.76-82, 88
	LIU R F , WANG J S , ZHANG H L , et al. Research progress and application of behavior tree technology[J]. Computer and Modernization, 2020, (2): 76- 82.76-82, 88
23	DRACHEN A , Thurau C , TOGELIUS J , et al. Game data mining[M]. London: Springer, Game Analytics, 2013: 205- 253.
24	BOYD J . A discourse on winning and losing[M].
25	中国科学院自动化研究所. 即时策略人机对抗平台[EB/OL]. [2020-11-1]. http://wargame.ia.ac.cn.
	Institute of Automation, Chinese Academy of Sciences. Man-machine confrontation platform[EB/OL]. [2020-11-1]. http://wargame.ia.ac.cn.

棋子类型	机动渗透阶段	中远攻击阶段	夺控阶段
红方步兵	步兵机动到夺控点10格内区域, 满足射击条件时进行射击	步兵机动到夺控点3格内, 满足引导射击时, 优先引导射击, 满足射击条件时进行射击, 发现敌人时为自行火炮提供目标矫正	对夺控点进行夺控, 满足条件时进行射击, 夺控完成后在夺控点1格范围内进行防守
红方装甲车	跟随步兵机动, 停在中远地域内的隐蔽射击点	在中远地域内找到射击点, 进行隐蔽的引导射击, 被敌人发现后机动到新的隐蔽点	若步兵未完成夺控, 则进行夺控, 夺控后在夺控点附近防守

AI单位	既定想定对抗胜率			临机想定对抗胜率
AI单位	执红胜率	执蓝胜率	总胜率	执红胜率	执蓝胜率	总胜率
1号AI	30.5	64	47.5	27.5	77.5	52.5
2号AI	44.5	73.5	59	40	74	57
3号AI	96	100	98	88	99	93.5
道·思-group	11	37	24	27.5	65.5	46.5
5号AI	5	38.5	21.5	0	1	0.5

[1]	李冬, 焦义文, 高泽夫, 杨文革, 毛飞龙, 滕飞. 基于GPU的相位干涉仪FX鉴相算法[J]. 系统工程与电子技术, 2022, 44(11): 3320-3329.
[2]	刘延钊, 黄志球, 沈国华, 王金永, 徐恒. 基于决策树和BN的自动驾驶车辆行为决策方法[J]. 系统工程与电子技术, 2022, 44(10): 3143-3154.
[3]	孙雪, 黄志球, 沈国华, 王金永, 徐恒. 基于本体和BN的无人车行为决策方法[J]. 系统工程与电子技术, 2021, 43(2): 452-465.
[4]	闫华, 汪贻生, 王锐淇, 刘波, 郭立卿, 肖骅. 基于GPU的大规模多阶段任务系统可靠性并行计算方法[J]. 系统工程与电子技术, 2019, 41(1): 215-222.
[5]	王宏伟，郑文秀. 计算DFT谱的改进递归算法[J]. 系统工程与电子技术, 2013, 35(11): 2263-2268.
[6]	钟何平, 唐劲松, 张学波. 基于最小不连续的分块相位解缠算法[J]. Journal of Systems Engineering and Electronics, 2012, 34(9): 1801-1806.
[7]	于荣欢, 吴玲达, 邓宝松, 瞿师. 基于ITM的复杂电磁环境并行计算方法[J]. Journal of Systems Engineering and Electronics, 2012, 34(7): 1339-1343.
[8]	李良超, 牛武斌, 吴振森. 空中复杂目标对背景红外辐射的散射的并行计算[J]. Journal of Systems Engineering and Electronics, 2011, 33(12): 2573-2576.
[9]	于荣欢, 吴玲达, 瞿师. 组网雷达探测能力的并行计算与可视化方法研究[J]. Journal of Systems Engineering and Electronics, 2011, 33(11): 2512-2516.
[10]	郭立新，麻军，王蕊，刘晓勇. MPI并行矩量法计算二维粗糙面波束电磁散射[J]. Journal of Systems Engineering and Electronics, 2010, 32(9): 1841-1845.
[11]	李妮1, 陈铮2, 龚光红1, 彭晓源1. 多核并行计算技术在景象匹配仿真中的应用[J]. Journal of Systems Engineering and Electronics, 2010, 32(2): 428-432.
[12]	张代远1,2. 基于分布式并行计算的神经网络算法[J]. Journal of Systems Engineering and Electronics, 2010, 32(2): 386-391.

群队级兵棋实体智能行为决策方法研究

Research on behavior decision-making of multi entities in group-level wargame

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 25

相关文章 12

编辑推荐

Metrics

本文评价