基于时序数据的火电厂主变压器绕组变形异常诊断方法

Time Sequence Data-based Method of Diagnosing Abnormal Deformation of Main Transformer Windings in Thermal Power Plants

通常在变压器停电后才可以对火电厂主变压器进行检测工作,导致异常运行的诊断结果的准确性受限,为此研究基于时序数据的火电厂主变压器绕组变形异常诊断方法。首先对主变压器绕组变形时序数据预处理,获得更精准的时序数据,接着根据时序数据信息选取主变压器绕组变形异常因素集,再通过输入主变压器绕组变形异常因素集定位变压器绕组变形异常,进而确定主变压器绕组变形故障程度,实现火电厂主变压器绕组变形异常诊断。实验结果表明该方法可准确诊断绕组变形异常,实际应用价值更高。

It is generally only after power outage can detection of main transformers in thermal power plants be carried out.This limits the accuracy of diagnostic results for abnormal operation.Therefore a method of diagnosing abnormal deformation of main transformer windings based on time sequence data was studied.First the deformation time sequence data of windings were preprocessed to obtain more accurate data.Second the set of abnormal deformation factors was determined according to the preprocessed data.Then the abnormal deformation was located by inputting the set of abnormal deformation factors,and the degree of deformation fault can be determined,achieving diagnosis of abnormal deformation.The experimental results show that this method can accurately diagnose abnormal winding deformation and has higher practical applicative value.