论文部分内容阅读
针对Bayes network(BN)能很好地表示变量的不确定性和相关性,并能直接进行不确定性推理的优势,提出了基于BN含失效节点的网络可靠性评估方法.建模方面,给出了通过节点分割去环构建有向无环图及条件概率表的具体方法;推理方面,通过节点消隐,减小BN规模,有效降低推理复杂度.分析了算法复杂度,并通过算例证明了算法的有效性和适用性.由于BN建模及推理的灵活性,该方法可研究同时含节点失效、共因失效的网络可靠性,况且通过BN推理除得到网络失效概率和部件重要度外,还可得到网络失效条件下任一部件失效或者共因失效发生的概率,为故障诊断和维护提供指导.
Aiming at Bayes network (BN), which can well represent the uncertainty and correlation of variables and directly take the advantage of uncertainty reasoning, a network reliability assessment method based on BN with failure nodes is proposed. The specific methods for constructing directed acyclic graph and conditional probability table are proposed by dividing the nodes into rings. In the reasoning, the nodal complexity is reduced by reducing the size of BN by node blanking, and the complexity of the algorithm is analyzed. The validity and applicability of the proposed algorithm are proved.Because of the flexibility of BN modeling and reasoning, the proposed method can be used to study the reliability of network with both node failure and common cause failure. Besides, BN failure probability and component importance In addition, the probability of failure or common failure of any component under the network failure condition can be obtained, providing guidance for fault diagnosis and maintenance.