Transformer partial discharge fault type identifying method and transformer partial discharge fault type identifying device

Abstract

The invention relates to a transformer internal discharge fault type identifying method and a transformer internal discharge fault type identifying device. The method comprises the steps of receiving transformer partial discharge ultrahigh-frequency signals collected by an external antenna sensor, extracting the characteristic parameters of the partial discharge ultrahigh-frequency signals, and inputting the characteristic parameters to a pre-trained classifier to get a transformer partial discharge fault type identification result. The time-frequency distribution map of the partial discharge ultrahigh-frequency signals is drawn and grayed, the initial characteristic parameters of time-frequency distribution of the partial discharge ultrahigh-frequency signals are extracted based on a gray level co-occurrence matrix, and a principal component analysis is made of the initial characteristic parameters to extract the characteristic parameters of the partial discharge ultrahigh-frequency signals. Characteristic parameters can be accurately extracted from the collected transformer partial discharge ultrahigh-frequency signals. The characteristic parameters are input to the pre-trained classifier, and the transformer internal discharge fault type can be identified quickly by the classifier. A basis is provided for transformer maintenance plan arrangement, and safe and stable operation of the power transformers is effectively guaranteed.

Description

technical field [0001] The invention relates to the field of fault diagnosis of power transformers, in particular to a method and device for identifying partial discharge fault types of transformers. Background technique [0002] Partial discharges occur when insulation defects occur inside power transformers. Different types of partial discharge have very different destructive effects on insulation, so it is necessary to distinguish various types of partial discharge. [0003] By analyzing the waveform of UHF signal generated by partial discharge, extracting its corresponding characteristic parameters and inputting them into the classifier, different types of insulation internal defects can be identified. The traditional method for extracting characteristic parameters of partial discharge UHF signals is mainly to perform wavelet decomposition and matrix decomposition on the signal, and extract the results of decomposition calculation as characteristic parameters. [0004]...

AI Technical Summary

Problems solved by technology

[0004] However, the above method can achieve high recognition accuracy when using the built-in antenna sensor, but when using the external antenna sensor for partial discharge detection, the recognition accuracy will drop due to the distortion of the UHF signal during propagation.

Images