基于多状态节点贝叶斯网络的HVDC系统可靠性评估

贝叶斯网络具有灵活的推理机制,能深刻地揭示系统内在的机理。该文提出了将贝叶斯网络应用于高压直流系统可靠性评估的方法。该方法采用多状态节点来描述系统及节点的多态性,有效地处理了特殊的备用情况,并提出了信念度矩阵来描述系统内部各层元件之间的连接关系,通过考查给定系统状态对应的信念度,可以评估系统中的薄弱环节。此外,利用贝叶斯网络的自动更新功能,还能根据当前的元件或子系统的状态对系统未来状态作出预测。最后,通过具体的算例对所提方法的有效性进行了分析和验证。

Reliability evaluation for HVDC systems based on Bayesian network with multi-state nodes

With flexible inference mechanisms, Bayesian networks can describe inherent relations of the system in depth. A method of reliability evaluation for HVDC systems based on Bayesian network is proposed in this paper. Muhi-state nodes are used to denote the multiple states of system and elements, which can effectively deal with the special backup modes. Belief matrices are proposed to describe the linking connection of elements from different layers, which can be used to evaluate the weak points of the system. Furthermore, automatic updating function of Bayesian networks can be used to forecast the future state of the system based on the current state of elements or sub-systems. The validity of the proposed method is proved by a test system.