A Partial Discharge Pattern Recognition System Applied to Power Equipment

Abstract

A partial discharge pattern recognition system applied to power equipment, including a signal preprocessing module that filters and amplifies the sensor coupling signal; then collects the signal processed by the signal preprocessing module, and then transmits the data to the data analysis module for further processing Analysis and processing, calculation of partial discharge PRPD spectrum and statistical characteristics, grouping the statistical characteristics of the discharge spectrum into the input layer of the BP network neural algorithm, obtaining the discharge results of different groups, and then comprehensively obtaining the discharge type according to the different weights assigned, Improved pattern recognition accuracy for partial discharges. The invention adopts high sampling rate and high repetition rate of pulse capture, which can improve the accuracy and sensitivity of the system.

Description

technical field [0001] The invention relates to the technical field of fault diagnosis of power equipment, in particular to a partial discharge pattern recognition system applied to power equipment. Background technique [0002] Partial discharge will occur in some weak parts of electrical equipment under the action of high field strength. Under certain conditions, partial discharge will lead to insulation degradation or even breakdown, endangering the safety of life and property. Therefore, live detection of partial discharge or Online monitoring can detect early defects or latent failures of equipment in time, which is of great significance to ensure the safety and stability of electrical equipment and power systems. There may be partial discharges caused by different types of defects inside the equipment, or the discharge signal is mixed with pulse interference, which requires pattern recognition of partial discharges to distinguish the degree of harm of different dischar...

AI Technical Summary

Problems solved by technology

[0005] (1) Statistical PRPD spectrum of partial discharge phase distribution is The spectrogram is triggered by a power frequency signal and continuously collected for 20ms. The spectrogram is calculated after multiple acquisitions. In the high sampling rate mode, the amount of collected data is large and the data analysis is slow

[0006] (2) In the phase analysis of the spectrogram, the traditional statistical features include skewness, steepness, peak number, etc. The statistical features have the same input weight in the partial discharge classifier, and there is no distinction between primary and secondary

Images