摘要
针对从“人在回路”的兵棋推演复盘数据中获取指挥员决策经验问题,提出了一种基于陆军合同战术兵棋复盘数据的火力打击行动决策要素关联分析模型,实现对其中武器使用规律的挖掘.该模型分为“数据兵棋要素关联挖掘”3个层级:数据层包括动态和静态数据;兵棋要素层构建了决策要素的定量化计算模型,实现原始数据向要素特征值的转化;关联挖掘层基于兵棋要素层的建模,提出了地形因素、协同规律和战果致因3个频繁项集模型构建方法,进而使用经典Apriori关联规则算法实现了对陆军战术对抗中武器使用规律的挖掘.将该模型应用于2018年第2届全国兵棋大赛复盘数据集,成功挖掘出坦克、战车、步兵作战单元在不同地形下的使用效能,以及不同作战单元之间的协同和武器对目标的打击效用规律.
In order to acquire commander’s decision-making experience from the“Man in loop”war-game replay data,a correlation analysis model of the fire striking elements is proposed to mine the rules of weapons usage from the replay of Army contract tactical war-game.The proposed model is composed of three layers:data layer,which stores both dynamic and static data of replay;elements layer,in which a quantitative computational model of fire attack elements is proposed to transform original replay data to features values;correlation mining layer,in which a frequent itemsets construction method is proposed to use three dimensions as terrain,cooperation and damage causality.And with these frequent itemsets,the classic Apriori algorithm can be applied to mine the rules of weapons usage in the Army contract tactical combat.Finally,the proposed model is applied to the dataset of the second national war-game competition.The experimental results successfully explored the e ectiveness of tanks,vehicles and infantry in di erent battle terrain,as well as the rule of coordination between di erent combat units and weapon strike e ectiveness against di erent targets.
作者
邢思远
倪晚成
张海东
闫科
XING Si-Yuan;NI Wan-Cheng;ZHANG Hai-Dong;YAN Ke(University of Chinese Academy of Sciences,Beijing 100049,China;Innovation Academy for Artificial Intelligence,Chinese Academy of Sciences,Beijing 100190,China;Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China;School of Joint Operation,National Defense University,Beijing 100091,China)
出处
《指挥与控制学报》
2020年第2期132-140,共9页
Journal of Command and Control
基金
国家自然科学基金(61906197)资助。
