基于数据矩阵化的电力计量装置黑盒故障研究

Black box failure of electric power metering device based on data matrix

由于电力计量装置黑盒故障诊断知识库中含有大量异常特征,导致故障诊断不准确、诊断时间长,由此提出基于数据策略矩阵化的电力计量装置黑盒故障诊断方法。构建由异常特征模型库和专家规则库组成的电力计量装置黑盒故障诊断知识库,利用异常特征模型库采集故障特征信息,通过专家规则库去获取异常特征编码与故障对应关系;在此基础上,采用基于局部采样数据信息的局部信息增量矩阵判断是否存在故障,完成电力计量装置黑盒故障诊断。实验结果表明:所研究方法在单机和集群两种情况下的电力计量装置黑盒故障诊断准确率均较高,诊断时间较短。

Due to a large number of abnormal features in the black box fault diagnosis knowledge base of power metering device,the fault diagnosis is not accurate,and the diagnosis time is long.Therefore,a black box fault diagnosis method of power metering device based on data strategy matrix is proposed.A black box fault diagnosis knowledge base is constructed of electric power metering device,which is composed of abnormal feature model base and expert rule base.The abnormal feature model base is used to collect fault feature information,and the corresponding relationship between abnormal feature code and fault is obtained through expert rule base.On this basis,the local information increment matrix based on local sampling data information is used to judge whether there is fault,and the electric power is completed Black box fault diagnosis of force measuring device.The experimental results show that the accuracy of black box fault diagnosis of the power metering device is high,and the diagnosis time is short under the condition of single machine and cluster.