基于EMEBN的空中目标战术意图识别方法  

Research on Air Target Tactical Intention Recognition Based on EMEBN

作　　者：杨雨田 杨杰 李建国 YANG Yutian;YANG Jie;LI Jianguo(North Automatic Control Technology Institute,Taiyuan 030006,China;The Third Military Representative Office in Nanjing,Nanjing 211153,China)

机构地区：[1]北方自动控制技术研究所,太原030006 [2]驻南京地区第三军事代表室,南京211153

出　　处：《火力与指挥控制》2022年第5期163-170,共8页Fire Control & Command Control

基　　金：“十三五”共性技术预研课题基金资助项目(41412010701)。

摘　　要：战术意图识别是战场态势分析与威胁评估中的重要一环,针对复杂战场环境中敌目标战术意图存在的动态性、序列性等问题,构建能够描述目标战术意图表达和推理的动态序列贝叶斯网络模型(dynamic series bayesian network,DSBN),分析多实体贝叶斯网络(multi-entities bayesian network,MEBN)在表达规则知识概率迁移关系和序列关系的局限性,提出基于扩展多实体贝叶斯网络(extended multi-entities bayesian network,EMEBN)的战术意图识别模型构建方法,最后通过实例验证该方法的可行性和有效性。Tactica l intention recognition is an important part of battlefield situation analysis and threat evaluation.Aiming at the dynamic and sequential problems of enemy target tactical intention in complex battlefield environment,the Dynamic Series Bayesian Network(DSBN) model is constructed to describe the expression and reasoning of tactical intention,the limitation of the Multi-Entities Bayesian Network(MEBN) in expressing probablistic transfer relation and sequence rule knowledge is analyzed,Then the construction method of tactical intention recognition model based on Extended Multi-Entities Bayesian Network(EMEBN) is proposed.The feasibility and effectiveness of this method is verified by a simulation example.

关 键 词：意图识别 动态序列贝叶斯 扩展多实体贝叶斯 规则知识 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]