Fault identification of high-voltage circuit breaker energy storage mechanism based on EEMD and GA-BP neural network

MA Li; HUO Yaojia; PAN Jie; LIU Huilan

doi:10.3969/j.issn.1672-3643.2026.03.005

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Fault identification of high-voltage circuit breaker energy storage mechanism based on EEMD and GA-BP neural network

更新时间：2026-10-10

- Fault identification of high-voltage circuit breaker energy storage mechanism based on EEMD and GA-BP neural network
- Ningxia Electric Power Issue 3, (2026)
- 作者机构：
  
  1. 国网宁夏电力有限公司电力科学研究院,宁夏,银川,750011
  2. 华北电力大学(保定),河北,保定,071003
- 作者简介：
- 基金信息：
- DOI：10.3969/j.issn.1672-3643.2026.03.005
  CLC：
- Online First：21 July 2026，
  
  Published：2026
- 稿件说明：
移动端阅览
MA Li, HUO Yaojia, PAN Jie, et al. Fault identification of high-voltage circuit breaker energy storage mechanism based on EEMD and GA-BP neural network[J]. Ningxia Electric Power, 2026, (3).
DOI：

MA Li, HUO Yaojia, PAN Jie, et al. Fault identification of high-voltage circuit breaker energy storage mechanism based on EEMD and GA-BP neural network[J]. Ningxia Electric Power, 2026, (3). DOI： 10.3969/j.issn.1672-3643.2026.03.005.

摘要

为实现高压断路器储能状态故障信号辨识，提出一种集合经验模态分解（ensemble empirical mode decomposition

EEMD）边际谱能量熵与遗传算法（genetic algorithm

GA）改进的误差反向传播（back propagation

BP）神经网络相结合的高压断路器储能状态辨识方法。首先采用EEMD对信号进行分解，得到一系列含有储能信号特征的本征模态函数（intrinsic mode function

IMF），利用相关系数筛选与原始信号相关的模态分量，提取前6阶IMF分量希尔伯特边际谱能量熵作为特征向量，最后，将构建的特征向量矩阵输入GA-BP进行分类训练，实验结果表明：基于EEMD边际谱能量熵与GA-BP相结合的故障辨识方法能够准确识别断路器储能机构故障。

Abstract

To achieve fault signal identification of the energy storage state of high-voltage circuit breakers

an identification method combining ensemble empirical mode decomposition (EEMD) marginal spectrum energy entropy with a genetic algorithm-optimized back propagation (GA-BP) neural network is proposed. First

the EEMD is adopted to decompose the signals

yielding a series of intrinsic mode functions (IMFs) that contain the characteristic information of energy storage signals. The correlation coefficient is then used to select the modal components correlated with the original signals

and the Hilbert marginal spectrum energy entropy of the first six IMF components is extracted as the feature vector. Finally

the constructed feature vector matrix is input into the GA-BP neural network for classification training. Experimental results show that the proposed fault identification method

combining EEMD marginal spectrum energy entropy with GA-BP

can accurately identify faults of the energy storage mechanism of circuit breakers.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Energy spectrum and energy entropy analysis on vibration signal of circuit breaker based on wavelet packet decomposition

Optimization and simulation on the covering air-supported membrane structure for a substation project

Wind power forecasting based on modal decomposition and TCN – transformer

A regional hierarchical energy storage planning method based on improved particle swarm optimization

Operational optimization of gas-electricity coupled systems considering distribution network power flow and high-penetration renewable energy

Related Author

WANG Lu

PENG Yanjun

PAN Ningbo

TENG Benke

CHEN Jianxi

Related Institution

No data

AI问答

⁰