基于MEBN的战术级空中目标意图识别

摘要

战术级空中目标意图识别是现代化防空作战中理解战场态势、预测目标行动的关键.构建空中目标意图识别模型的核心问题是如何表示不确定的意图因素以及它们之间的关系.分析了具有典型防空作战意义的想定,并结合军事经验总结了影响意图识别的因素.提出了应用多实体贝叶斯网络(Multi-Entities Bayesian Network)描述空中目标意图,在贝叶斯推理中融入逻辑理论,采用具体的基于知识的SSBN构建算法动态地进行空中目标意图识别.通过仿真实验,验证了此方法的有效性和可行性.%In the modern air-defense battle, tactical air target intention recognition is the key point of comprehending the battlefield situation and predicting the coming air target action. How to represent the uncertain intention factors and the relationship among them is the fundamental issue. Through analyzing a typical air-defense battle scenario and combining the military experiences, the paper summaries the influence factors of intention recognition. The paper presents the MEBN to describe the air target intentions,blends the logic theory into the Bayesia theory, and dynamically implement air target intention recognition by knowledge-based SSBN constructing algorithm. The simulation experiment verifies the validity and the feasibility of the method.