Transformer Winding Fault Diagnosis Method based on Frequency Response Multilevel

Autotransformers are one of the key devices in the power supply system of high-speed railroads. The existing FRA models and diagnostic methods are difficult to recognize the deformation faults occurring in the windings. This paper proposes a FRA curve interpretation method based on multilevel decomposition to obtain more information about the frequency response curve, analyze the small changes of the curve, and carry out shunt capacitance simulation tests on different windings and locations of the autotransformer to verify and analyze the results. The results show that the FRA curve after multilevel decomposition has strong correlation with the original frequency response curve, and there are differences in the curve changes under different faults after transforming it into a polar plot, and the characteristic correlation values of different fault windings are obviously different, and different faults have their corresponding laws. According to the 12 image features extracted from the polar plot after 2-level decomposition and the corresponding feature correlation, different fault windings, fault locations and fault degree can be distinguished in the case of shunt capacitor fault of autotransformer.