Fault Classification Method of Motor Based on Variational Modal Decomposition and CNN-BiGRU-Attention Neural Network

SI Chengzhi; HUI Shixian; XING Chao; DENG Can

doi:10.13296/j.1001-1609.hva.2025.05.020

您当前的位置：

首页 >

文章列表页 >

Fault Classification Method of Motor Based on Variational Modal Decomposition and CNN-BiGRU-Attention Neural Network

The Key Technologies for Energy Conservation and Carbon Reduction in New Power Systems | 更新时间：2025-05-25

- Fault Classification Method of Motor Based on Variational Modal Decomposition and CNN-BiGRU-Attention Neural Network
- High Voltage Apparatus Vol. 61, Issue 5, Pages: 189-196(2025)
- 作者机构：
  
  1.云南电网有限责任公司文山供电局，云南　文山　 663000
  2.云南电网有限责任公司电力科学研究院，昆明　 650217
- 作者简介：
- 基金信息：
- DOI：10.13296/j.1001-1609.hva.2025.05.020
  CLC：
- Received：13 December 2024，
  
  Revised：21 February 2025，
  
  Published：16 May 2025
- 稿件说明：
移动端阅览
SI Chengzhi, HUI Shixian, XING Chao, et al. Fault Classification Method of Motor Based on Variational Modal Decomposition and CNN-BiGRU-Attention Neural Network[J]. High voltage apparatus, 2025, 61(5): 189-196.
DOI：

SI Chengzhi, HUI Shixian, XING Chao, et al. Fault Classification Method of Motor Based on Variational Modal Decomposition and CNN-BiGRU-Attention Neural Network[J]. High voltage apparatus, 2025, 61(5): 189-196. DOI： 10.13296/j.1001-1609.hva.2025.05.020.

摘要

为了解决电机故障分类结果准确性较低的问题，提出一种基于变分模态分解（variational mode decomposition，VMD）与CNN-BiGRU-Attention神经网络模型相结合的电机故障诊断方法。首先，对凯斯西储大学电机数据进行预处理，然后利用VMD方法将复杂的电机信号分解成多个IMF（intrinsic mode function）分量，构建电机故障特征向量。最后，建立特征训练CNN-BiGRU-Attention神经网络模型，基于模型的迭代训练，诊断电机故障的具体类型。实验结果表明，提出的诊断方法应用后，电机故障分类错误诊断数量较少，测试集准确率达到97%以上，方法具有较高的鲁棒性和精确性。

Abstract

For solving the problem of low accuracy of fault classification results of motor

a kind of fault classification method for motor based on variational mode decomposition (VMD) and CNN-BiGRU-Attention neural network model is proposed. Firstly

the motor data of Case Western Reserve University is preprocessed

and then the complex motor signal is decomposed into multiple intrinsic mode function (IMF) components by VMD method and the fault feature vector of motor is constructed. Finally

the feature training CNN-BiGRU-Attention neural network model is set up

and the specific types of faults of motor are diagnosed based on the iterative training of the model. The experimental results show that after the application of the proposed diagnosis method

the number fault classification errors of motor is less and the accuracy of test set is up to 97%. The method has high robustness and accuracy.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Acoustic and Vibration Joint Fault Diagnosis Method of High Voltage Circuit Breaker Based on VMD and KFCM-SVM

Failure Statistics and Countermeasures of UHVDC Converter Stations

Related Author

HUO Yaojia

WU Yang

CHANG Jing

HAN Li

QIAN Yong

FANG Jizhong

MA Li

CHENG Jiandeng

Related Institution

Ningxia Electric Power Research Institute, State Grid Ningxia Electric Power Company

Kunming Bureau of CSG EHV Transmission Company

South China University of Technology

⁰