A method and device for extracting characteristic bands of partial discharge spectrum based on dimension reduction

Abstract

The present application discloses a method and device for extracting characteristic bands of partial discharge spectrum based on dimensionality reduction. The method includes: using a multispectral partial discharge detection method to collect partial discharge multispectral signals at different voltages; The optical pulse amplitude and phase information of the partial discharge multispectral signal in each spectral band under the voltage of the group; the average value of the amplitude of the optical pulse collected in each spectral band of Calculate the average value of the amplitude of the internal light pulse, and normalize the data respectively to obtain the ratio of the average optical radiation intensity in each spectral band, so that the ratio of the average optical radiation intensity in each spectral band is unchanged and the sum is 1; The standard deviation of the ratio of the average optical radiation intensity in each spectral band in each group of voltages; according to actual requirements, a set number of spectral bands with a larger standard deviation are selected as the characteristic band under the partial discharge discharge type.

Description

technical field [0001] The present application relates to the technical field of power equipment state monitoring and fault diagnosis, and in particular, to a method and device for extracting characteristic bands of partial discharge spectra based on dimension reduction. Background technique [0002] As an effective method to find insulation defects, partial discharge detection of power equipment plays an important role in the diagnosis of power equipment. Usually partial discharge is accompanied by the physical phenomenon of optical radiation, so the development of partial discharge can be laterally reflected by detecting the optical radiation signal. The partial discharge spectrum can reflect the microscopic information such as electron temperature, excitation cross section and development mode, and the spectral characteristics can be used to conduct in-depth analysis of the discharge mechanism and the degree of insulation degradation. Combining discharge statistical info...

AI Technical Summary

Problems solved by technology

[0004] This application provides a method and device for extracting characteristic bands of partial discharge spectra based on dimensionality reduction, to solve the problem that the existing partial discharge diagnosis technology of electric power equipment has a huge amount of data in the process of collecting data, which leads to a decrease in the operation and processing efficiency of the computer question

Images