基于混合贝叶斯网的空域目标威胁评估方法

针对空域目标威胁评估既需要综合考虑离散型变量和连续型变量影响，又需要具有不确定性推理能力的特点，建立了一种基于混合贝叶斯网的空域目标威胁评估模型。提出了结合主观经验和客观历史数据进行连续型网络参数学习的方法，提高了决策模型定量描述问题域中变量间依赖关系的准确性。运用团树传播算法进行空域目标威胁评估模型推理，仿真结果验证了该方法的有效性。

Threat assessment of aerial targets based on hybrid Bayesian network

Aiming at the characteristic that a threat assessment of aerial targets requires to have the ability of an uncertain inference, as well as consider the influence of both discrete variables and continuous variables comprehensively, a Bayesian network model for a threat assessment of aerial targets is constructed. The learning method of continuous parameters with subjective knowledge and historical data is proposed to promote the accuracy for describing the dependency relationship among variables. The clique tree algorithm is used to make an inference of the threat assessment model of aerial targets. Simulation results verify the validity of the proposed method.