基于动态贝叶斯网络的无人战车目标威胁评估

Threat Assessment of Unmanned Combat Vehicle Target Based on Dynamic Bayesian Network

针对传统地面目标威胁评估方法考虑目标类型单一、易受人为主观因素影响且多为静态评估的问题,提出一种基于动态贝叶斯网络的无人战车目标威胁评估方法。分析了无人战车作战问题,选取合理的目标特征并进行模糊处理;根据选取的目标特征,结合专家知识,建立了威胁评估的静态贝叶斯网络;基于动态贝叶斯网络理论,将已建立的静态贝叶斯网络扩展为动态贝叶斯网络;最后,结合算例进行了仿真,并将动态评估结果与静态结果进行对比,表明基于动态贝叶斯网络的威胁评估准确率高、鲁棒性强,更适用于高动态强对抗的实际战场环境。

Aiming at the problems that the traditional ground target threat assessment method considers single type of targets,easily affected by human subjective factors and mostly static assessment,an unmanned combat vehicle target threat assessment method based on Dynamic Bayesian network is proposed.Firstly,the combat problems of unmanned combat vehicle are analyzed,and the reasonable target features are selected and fuzzy processing is carried out.Secondly,according to the selected target features and expert knowledge,the static Bayesian network for threat assessment is established.Then,based on the dynamic Bayesian network theory,the established static Bayesian network is extended to a dynamic Bayesian one.Finally,a numerical example is simulated.The results of dynamic and static assessments are made comparision.The results show that the dynamic threat assessment based on Bayesian network is more accurate robust and more applicable to the actual battlefield.