Power distribution network fault positioning method based on wavelet transformation and CNN

Abstract

The invention relates to the power distribution network fault positioning technology field and particularly relates to a power distribution network fault positioning method based on wavelet transformation and the CNN. Wavelet transform multi-scale analysis is utilized to decompose fault current data, in the parallel coordinate system, modulus maxima are calibrated in order, points are connected insequence to form a line graph, and lastly, the line graph is processed to be a grayscale image used as the input of the CNN, the powerful feature extraction capability of the CNN is utilized to extract hidden topological structure features in the data, a machine is enabled to automatically identify the traveling wave head, and the B-type traveling wave ranging method is utilized to realize faultpositioning. The method is advantaged in that shortcomings of not-enough wavelet transform high-scale time resolution, large low-scale noise interference and easy false determination of the travelingwave head are overcome, the parallel coordinate system is utilized to fully combine the characteristics of wavelet transform and the convolutional neural network to achieve high-scale to low-scale automatic search for the traveling wave head, and strong anti-interference ability and high accuracy are achieved.

Description

technical field [0001] The invention relates to the technical field of distribution network fault location, and more specifically, relates to a distribution network fault location method based on wavelet transform and CNN. Background technique [0002] Short-circuit faults in the distribution network will endanger the safe and stable operation of the system and cause unnecessary economic losses. Fast and accurate fault location methods are conducive to troubleshooting and improve the operational reliability of the system. At present, the distribution network fault location is mainly divided into fault section location and fault precise location from the function. The location of the fault section is mainly to judge the fault branch, which cannot fully satisfy the rapid troubleshooting of the fault; the precise fault location methods are mainly the impedance method and the traveling wave method. The impedance method is greatly affected by the transition resistance and the loc...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the above technical problems, the technical solution adopted in the present invention is: mainly aiming at the shortcomings of the existing distribution network traveling wave positioning method that the machine automatically detects the head of the traveling wave is not accurate, and manual intervention is required to capture the head of the wave, a multi-scale method combined with wavelet is proposed. Analyze and use the feature extraction ability of CNN to automatically identify traveling wave heads

Images