Method for enhancing partial discharge signal PRPD atlas data based on CGAN network

Abstract

According to the method for enhancing the partial discharge signal PRPD atlas data based on the CGAN network provided by the invention, opencv is adopted to preprocess an original image, redundant information in the image is eliminated, useful information is recovered, the detectability of the useful information is enhanced, and the reliability of the information is improved; the new data is generated by utilizing the CGAN network, the generation of the new data is based on the original data and is similar to the original data, the quality of the new data is ensured, the original data volume is enriched, the problems of insufficient data and relatively low data quality are solved, and the purpose of enhancing the partial discharge PRPD atlas data is achieved.

Description

technical field [0001] The invention relates to the technical field of high-voltage electrical equipment insulation state detection, in particular to a method for enhancing partial discharge signal PRPD map data based on a CGAN network. Background technique [0002] The UHF method is widely used in the detection of partial discharge, and the pattern recognition of partial discharge can be carried out based on the PRPD spectrum of the UHF signal. The algorithm model of pattern recognition is mainly designed based on machine learning methods. The premise of this type of algorithm is that a large amount of sample data is required to train the algorithm model. With the development of deep learning in recent years, deep neural networks have made revolutionary breakthroughs in classification tasks, and they have also made continuous breakthroughs in the direction of partial discharge pattern recognition. The classifier of the deep neural network can achieve high accuracy under th...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the deficiencies in the prior art, the present invention provides a method for enhancing the partial discharge signal PRPD map data based on the CGAN network, generating new data close to the original data based on the CGAN network, ensuring the quality of the new data, and enriching the original data volume , solve the problem of insufficient data and low data quality, and achieve the purpose of enhancing PRPD map data

Images