CNN (Convolutional Neural Network) based partial discharge fault diagnosis method of power equipment

Abstract

The invention discloses a CNN based partial discharge fault diagnosis method of power equipment, and relates to the field of electric variable measurement. The method comprises the following steps that partial discharge signals of a typical fault defect model are collected, and a PRPD atlas database is established; a grayscale map construction module obtains a partial discharge grayscale map via PRPD; a feature extraction module which is established by using the CNN with a residual error structure extracts identification features from the grayscale map; the identification features are sent toa classifier to carry out identification training; and the grayscale map construction module, the feature extraction module and the classifier serve as a diagnosis module of a detection instrument. According to the method, RPPD data is converted into the grayscale map, the identifying features of the grayscale map are extracted via the strong feature adaptive extraction capability of the CNN withthe residual error structure, the features are applied to the classic classifier, a deep learning method is combined with a traditional machine learning method effectively, the extracted features arehigh in identification degree, and the accuracy of fault diagnosis can be improved.

Description

technical field [0001] The present invention relates to the field of shared technology, and more specifically refers to a method for diagnosing partial discharge faults of power equipment based on convolutional neural networks. Background technique [0002] During the design, manufacture, transportation, installation, long-term operation and maintenance of power equipment, various latent insulation defects that may cause serious harm may appear inside the equipment, resulting in different types of partial discharge (PD). Different partial discharge types reflect different insulation degradation mechanisms and different degrees of damage to the insulation capacity of equipment. Therefore, it is of great significance to carry out research on the characteristics of partial discharge of power equipment and its diagnostic methods for the safe and stable operation of power equipment and the guidance of on-site fault handling. [0003] For the identification of partial discharge si...

AI Technical Summary

Problems solved by technology

This technology has the following disadvantages: (1) The large amount of data requires the device to store and transmit a large amount of data, which affects the working efficiency of the device; (2) It is impossible to collect pulses with a frequency of GHz in partial discharge

Images