基于贝叶斯网络的态势评估

Situation Assessment Based on Bayesian Networks

作为一种知识表示和进行概率推理的框架,贝叶斯网络在具有内在不确定性的推理和决策问题中得到了广泛的应用。因果推理是态势评估中的一个重要环节,用贝叶斯网络找出态势假设和事件之间的潜在关系,正是态势评估所需完成的功能。根据态势与事件之间不同的连接关系建立了态势评估的贝叶斯网络模型,并分别介绍了相应的信息传播算法,最后以一个实例来说明该网络的计算过程。

As a knowledge representation framework and a kind of probability inference engine, Bayesian networks are widely used in reasoning and decision-making. Causal reasoning plays an important role in situation assessment (SA). Finding out the hidden patterns between situation hypothesis and events is the function needed in situation assessment. In this paper, a Bayesian network model for situation assessment is set up. Based on different link relationship, the fusion and propagation algorithms of new information through the network are introduced. Finally, an example is provided to show the calculation process of the network.