基于贝叶斯网络和故障区域识别电网故障诊断的研究.pdfVIP

第 38 卷 第 12 期 电力系统保护与控制 Vol.38 No.12 2010 年 6 月 16 日 Power System Protection and Control Jun.16, 2010 基于贝叶斯网络和故障区域识别的电网故障诊断研究 何小飞，童晓阳，周 曙 （西南交通大学电气工程学院，四川 成都 610031） 摘要：针对电网中因元件故障导致多个保护连锁动作的现象，提出了基于故障区域识别和贝叶斯网络的新型电网故障诊断方 法。以电网元件为根节点分别建立贝叶斯网络，利用实时接线分析法确定故障区域以缩小诊断范围，再对故障区域中各元件 的贝叶斯网络，利用反向推理得到各元件故障概率。结合SCADA/RMS系统信息，正向推理和相应计算推断出故障元件、拒动 与误动的保护装置及断路器。算例的仿真结果证明了该方法的有效性，对于级联跳闸情况可准确诊断出故障。 关键词：电网故障诊断；贝叶斯网络；故障区域识别；SCADA/RMS系统；级联跳闸 Power system fault diagnosis based on Bayesian network and fault section location HE Xiao-fei ，TONG Xiao-yang ，ZHOU Shu （Electrical Engineering Institute ，Southwest Jiaotong University，Chengdu 610031，China ） Abstract ：To solve the problem of a number of components failure causing cascading protection actions，a new type of fault diagnosis method based on Bayesian network and fault identification is studied ．First of all ，the root nodes of Bayesian network based on the grid components are set up separately, a real-time analysis method is used to determine wiring fault diagnosis to narrow the scope of fault zones，and then，for the Bayesian network components in the fault zones，the reverse inference failure probability of all components is made ．Combining with the system information from SCADA ／RMS （relay management system ），the forward reasoning and analysis calculation are made to infer the probability of maloperation and misoperation of protection and breaker ．The simulation testing illustrates the reliability and effectiveness of the method and the method can make accurate diagnosis for cascading tripping ． Key words ：power system fault diagnosis ；Bayesian network ；fault section location ；SCADA/RMS system；cascading trip 中图分类号： TM734 文献标识码：A 文章编号： 1674-341