Fault identification based on BP neural network and wavelet packet in power systems

This paper proposes a fault identification method based on BP neural network and wavelet packet, which extracts the fault transient eigenvalues of the three-phase current and zero-sequence current from measurement data under the fault conditions. Firstly, the eigenvalues of three-phasors are sampled...

Full description

Saved in:
Bibliographic Details
Published inIOP conference series. Earth and environmental science Vol. 645; no. 1; pp. 12057 - 12070
Main Authors Tao, Fang, Qian, Sun, Hangli, Jian, Ning, Li, Yan, Zhou, Yanjie, She, Zhangao, Li, Ying, Cai, Leiyu, Zhao, Jiang, Li
Format Journal Article
LanguageEnglish
Published Bristol IOP Publishing 01.01.2021
Subjects
Circuits
Eigenvalues
Identification methods
Neural networks
Phase current
Phasors
Short circuits
Wavelet transforms
Zero sequence current
Online AccessGet full text

Cover

More Information
Summary:This paper proposes a fault identification method based on BP neural network and wavelet packet, which extracts the fault transient eigenvalues of the three-phase current and zero-sequence current from measurement data under the fault conditions. Firstly, the eigenvalues of three-phasors are sampled under the typical faults, such as single-phase ground fault, two-phase short-circuit fault, two-phase ground short-circuits fault, and three-phase short-circuit fault. Secondly, the three-phase current and the zero-sequence current are subjected to wavelet packet transform to extract the eigenvalues, which are viewed as the input of the neural network to determine the fault type. Finally, simulation results show that the proposed method can give reliable identification results under different fault conditions.
ISSN:1755-1307
1755-1315
DOI:10.1088/1755-1315/645/1/012057