基于贝叶斯网络和故障录波数据的电网故障综合诊断方法

A Comprehensive Fault Diagnosis Method Based on Bayesian Network and Fault Recorder Data

电网故障诊断是保证系统安全运行的基础,故障录波系统提供的信息为电网故障的精确诊断提供了重要依据。为了有效利用故障录波数据,并在信息缺失或不确定条件下精确诊断电网故障,研究了基于贝叶斯网络和故障录波数据的电网故障综合诊断方法。通过在故障录波联网系统的主站建立输电网贝叶斯模型,结合从故障录波系统得到的模拟量和开关量数据,从而构成基于贝叶斯网络的电网故障识别模型。将贝叶斯网络得出的高可信度诊断结果融入专家系统的知识库,形成完备的故障诊断专家系统。实例分析表明,可快速并准确地定位故障元件,提高电网故障处理效率。

The data from fault recorder network are not utilized thoroughly at present. In order to overcome the disadvantages of data acquisition of power grid fault diagnosis and the defects of expert system which can not get the valid conclusion under the condition of incomplete information, the work mechanism of Bayesian network is studied by modeling a Bayesian network on the sever of fault recorder network in this paper. By using digital and analog data from fault recorder system, the diagnosis results with a high credibility are input into the acknowledge data base of expert system, which helped build more complete expert system. The result of practical fault diagnosis examples demonstrates the method mentioned above can locate the fault device fast and accurately.