Voltage sag source identification method based on wavelet transform

Abstract

The invention discloses a voltage sag source identification method based on wavelet transform. The method comprises the following steps: 1) carrying out effective value transformation on three-phase voltage data, and determining voltage grade and sampling frequency; 2) determining a sag start point, a sag recovery point and a sag point range; 3) preliminarily judging a voltage sag source by utilizing distribution characteristics of voltage data within the sag point range; and 4) when the preliminary judgment cannot draw the conclusion, carrying out wavelet transform on the three-phase voltagedata, and then, further judging the voltage sag source by utilizing wavelet transform values of the sag start point and the sag recovery point of each phase of voltage and relationship between the wavelet transform values. The method can reliably find the sag source when the voltage has sag; and meanwhile, the method can realize full machine identification based on three-phase voltage sample data,can greatly improve identification efficiency of the voltage sag source and provide more reliable and efficient data basis for maintenance and operation of an electrical power system.

Description

technical field [0001] The invention relates to the technical field of voltage sag identification, in particular to a wavelet transform-based identification method for a voltage sag source. Background technique [0002] At present, the usual identification method of voltage sag source is: use the voltage waveform collected by the voltage sag recording device, use big data clustering to cluster and identify the sag data, and then manually label each category to judge. Its shortcomings are: the selection of clustering feature values ​​is mostly based on empirical judgments, and there is no clustering result supported by theory, resulting in inconsistent clustering results for different data sources; at the same time, the method requires manual judgment to label each cluster, which cannot be done completely. machine recognition. Contents of the invention [0003] The object of the present invention is to provide a wavelet transform-based identification method for voltage sa...

AI Technical Summary

Problems solved by technology

Its shortcomings are: the selection of clustering feature values ​​is mostly empirical judgment, and there is no clustering result supported by theory, which leads to inconsistent clustering results for different data sources; at the same time, the method requires manual judgment to label each cluster, which cannot be done completely. machine recognition

Images