Research On Transformer High Voltage Oscillating Wave Test And Winding Fault Diagnosis

In recent years,the scale of the power transmission system is expanding,and it is developing in the direction of ultra-high voltage,large capacity,and national interconnection.Electrical equipment is an element of the power transmission system and the basis for ensuring the safety of the power transmission system.As one of the most important and valuable electrical equipment,the safety of the transformer operation is closely related to the stability of the power transmission system.In daily operation,the transformer may face extreme conditions such as high circuit and overvoltage,which causes the winding to suffer from a greater impact,and the risk of winding failure increases significantly.Therefore,it is necessary to complete the fault diagnosis of transformer winding.The transformer state assessment is conducted during maintenance time for regular inspection.Frequency response analysis and power frequency tests have been widely used.Due to the rapid increase in the number of transformers,the existing methods have the following problems: there are many test items for a single transformer,and the requirements of different test items are different.Various maintenance needs to replace cables and instruments.A complete overhaul for a large transformer group requires a long period of outage.However,modern power systems allow very limited outage overhaul time.Therefore,proposing an efficient and accurate transformer winding test method can effectively improve the maintenance efficiency,reduce the maintenance cost,find potential faults,and eliminate potential safety hazards,which is very important for the stable operation of the transformer and even the power system.To improve the maintenance efficiency and accuracy of transformer winding state assessment,this paper proposes the transformer high-voltage oscillation wave test and winding fault diagnosis method.The main work of this paper includes:1)Based on the time domain and frequency domain,an analysis is conducted of the oscillation wave test principle.The high-voltage oscillating wave test is carried out to verify the feasibility.The combination of the theoretical analysis and field test helps to select the optimal excitation parameters and the connection.2)Combined with the theoretical derivation of parameters and state-space equations,the oscillating wave model of the transformer is constructed.Based on the oscillating wave signal time-frequency distribution,an optimized method for time-varying parameters is proposed,which effectively improves the accuracy of the model.Based on the test platform and electromagnetic model,typical winding faults are simulated and the evolution law of oscillatory waves is analyzed.3)Based on the rules of the frequency-domain oscillating wave signal,the intelligent sub-band method and multi-resolution segmentation method are proposed for frequency domain curves.The combination of multi-scale fusion features and kernel extreme learning machine completes the intelligent fault location of the winding.The application on the actual transformer cases verified the feasibility of the locating method.In comparison with the traditional method,the oscillating wave method presents advantages in high efficiency.4)Based on the rules of the time-domain oscillating wave signal,the graph transformation method and time-frequency transformation method are proposed.The regional features and color features are extracted from the W-polar diagram and time-frequency distribution diagram respectively.By the intelligent process of the extracted features,the centroid and color moment are chosen for classification and degree evaluation.With the support vector machine,the automatic fault diagnosis is completed.The application on the actual transformer cases verified the feasibility of the fault diagnosis method.On the basis of this paper,it can be seen that the high-voltage oscillating wave method can accurately evaluate the mechanical structure and insulation state of transformer windings.Furthermore,it takes only 15 minutes,and maintenance efficiency is greatly improved.