基于贝叶斯网络的分布式输电网故障诊断

Distributed Fault Diagnosis in Power System Based on Bayesian Networks

针对输电网拓扑结构日渐复杂且发生故障后报警信息存在不确定性的情况,提出了一种基于贝叶斯网络的分布式输电网故障诊断方法。分别建立了线路、变压器、母线的分布式元件诊断模型和联合诊断模型,并对其进行了分层。结合SCADA/RMS系统信息,对元件诊断模型采用后向推理,确定可疑故障元件;再由联合诊断模型推理,得到具有最大后验概率的故障元件组合。经过算例仿真,验证了该方法的正确性和有效性。

According to the complicated topology of transmission grid and the uncertainties of the alarm information af-ter fault, this article proposes a distributed power grid fault diagnosis method based on Bayesian networks. The distrib-uted element diagnosis model and combined diagnostic model of line, transformer and bus were constructed and de-laminated. The suspicious faulty components were inferred from the element diagnosis model that combined with the SCADA/RMS （relay management system） system message. Then, through the reasoning of combined diagnosis mod-el, the faulty components of power system were achieved, which fit the maximum posterior probability. Calculation ex-amples prove that the proposed method is correct and available.