基于直觉模糊贝叶斯网络的不确定性推理方法

摘要

针对模糊贝叶斯网络模型对战场态势信息不确定性描述与推理方面的局限性，利用直觉模糊属性函数表示贝叶斯网络节点概率以及变量间的因果依赖关系，给出了直觉模糊贝叶斯网络（IFBN）的定义，证明了节点直觉模糊性在贝叶斯网络推理中的可传播性，并提出了基于IFBN的不确定性推理方法，有效地提高了贝叶斯网络模型推理的准确性，增强了军事态势评估系统形成正确战场感知的作战效能。最后通过典型实例验证了该方法的有效性和优越性。%To overcome the constraint of Fuzzy Bayesian Network on description and reasoning for complex and uncertain time information of battle situation,the conception of Intuitionistic Fuzzy Bayesian Network,which uses the intuitionistic fuzzy property function to express the probability of the Bayesian Network nodes and the causal dependency of the variables,is proposed. And the node’s intuitionistic fuzzy transmissibility in Bayesian Network is proved. Meanwhile,the uncertain time reasoning based on IFTBN is advanced,promoting the accuracy of the Bayesian Network reasoning and enhancing the performance of shaping the proper battle sense in military situation assessment system. The typical example validates the presented approach’s efficiency and its preference.