基于动态云贝叶斯网络的舰艇防空目标威胁评估

针对目标威胁评估中信息表达的不确定性以及威胁评估模型专家网络结构的主观性,提出一种基于结构学习的动态云贝叶斯网络评估模型。首先,利用云模型良好的知识表达能力定量描述不确定连续性信息;其次,使用爬山算法进行结构学习,综合专家提出的网络结构构建贝叶斯网络;接着引入时间变量,将其扩展成为动态贝叶斯网络,然后用最大似然概率估计算法学习网络参数;最后,利用联合树算法对动态云贝叶斯网络进行推理评估。仿真结果表明,该模型能够有效的对观测信息进行威胁评估,具有合理性和可行性。

Threat Assessment of Ship Air Defense Target Based on Dynamic Cloud Bayesian Network

Aiming at the uncertainty of information expression in target threat assessment and the subjectivity of expert network structure in threat assessment model,a dynamic cloud Bayesian network assessment model based on structure learning is proposed.Firstly,this paper uses the cloud model with good knowledge expression ability to describe the uncertain continuous information quantitatively;secondly,uses the Hill Climbing algorithm to perform structure learning,and uses the network structure proposed by experts to construct Bayesian network,and introduces the time variable to expand it to dynamic Bayesian network,then uses the maximum likelihood estimation algorithm to learn the network parameter,finally uses the junction tree algorithm to evaluate the dynamic cloud Bayesian network.The simulation results show that the model can effectively evaluate the observed information and is reasonable and feasible.