Research On On-line Monitoring And Fault Diagnosis Of Partial Discharge In Hydropower Units

Hydropower unit is a core equipment of the hydropower station,which plays an important role in the station and the whole power system.Whose safety and reliability are closely related to the operation condition of the whole system.Furthermore,in the process of actual operation,power transformer is unavoidable to subject to such outside factor influence as electricity,machinery and heat,and so forth further causing its winding insulation deterioration to produce partial discharge(PD)phenomena,threatening the safety of operation of the whole system.Therefore,it is essential to monitor the insulation conditionand provide a proper maintenance action for in-service hydropower units.In this paper,taking hydropower units as an example,the aging mechanism of the main insulation system,partial discharge fault diagnosis and fault evaluation are studied.A large number of noise signals are usually contained in the partial discharge signal of the hydroelectric unit.The traditional wavelet transform method can suppress the noise interference to a certain extent,but there is a problem of misjudgement and leakage.Therefore,this paper introduces the theory of correlation space domain method,and applies the concept of quantile to each of the wavelet transform.A signal detection method based on the correlation probability wavelet transform is proposed to effectively suppress the noise interference in the signal.PD signal is of strong nonlinearity and time variation.And in the in situ detection,it is frequently subject to the overlap of many interference signals.Which leads great significant to signals feature extraction and fault diagnosis.Therefore,how to accurately extract signals feature is the key to PD faults identification.So this paper introduces manifold learning theory to partial discharge signals processing,and suggests a novel method for the PD signal feature extraction based on manifold learning theory.This method can effectively remove the background noise,solved the partial discharge signal frequency dispersion,denoising difficulties and other issues.The hydropower units partial discharge phenomenon is divided into three typical insulation defects.Then fault diagnosis system of hydropower units based on SVM-KNN algorithm is established,taking the partial discharge signals of time-frequency characteristic parameters as input vectors.The result indicates that this method is able to identify and classify different partial discharge faults.According to the practical operation of hydropower units and discharge energy,the generator partial discharge fault degree is divided into three state level,normal,abnormal and warning.N+and N-,umax+ and umax-,Qc are selected to describe fault degree.And the fuzzy comprehensive evaluation model is established,which can realize the fuzzy evaluation of the severity of the unit.Above all,it shows that methods of hydropower units partial discharge feature extraction,the fault pattern recognition and fault evaluation proposed in this paper are obtained has important academic significance and engineering application value.