Comprehensive Fault Diagnosis of Distribution Network Based on Four Axis Characteristics of "Time, Signal, Algorithm and Sample"

Abstract

With the increasing demand for reliable power supply of distribution network in economic society, how to realize sensitive and reliable perception of holographic information of distribution network and accurate fault diagnosis of urban distribution network has become the key to realize intelligent distribution network. Aiming at the problem of sensitive sensing and accurate positioning of distribution network, this paper proposes a comprehensive fault diagnosis method for distribution network based on signal, algorithm and sample. The fault features of time, signal, algorithm and sample are used for accurate analysis. The multi-level fault analysis method of switch, feeder and substation is used to build a multi-dimensional fault feature analysis library, and a comprehensive fault diagnosis model of distribution network is established, which integrates data fusion, feature extraction and intelligent decision-making. Using waveform correlation, empirical mode decomposition technology and other fault feature extraction methods, as well as the introduction of depth of historical sample analysis method, a multi-objective normalized distribution network comprehensive fault diagnosis method is established, which provides an effective means for accurate fault identification of smart distribution network and improves power supply reliability.