LOW VOLTAGE APPARATUS ›› 2024, Vol. 0 ›› Issue (10): 7-18.doi: 10.16628/j.cnki.2095-8188.2024.10.002

• Research & Analysis • Previous Articles     Next Articles

Electrical Life Prediction Method of DC Circuit Breaker Based on Residual Network Regression

YANG Nuoming1, WU Ziran1,2, LIU Zezhou3, WU Chenxi1, WU Guichu4, LIN Yigang1,2   

  1. 1. Low Voltage Apparatus Technology Research Center of Zhejiang(Wenzhou University), Wenzhou 325035, China
    2. Technology Institute of Wenzhou University in Yueqing, Wenzhou 325699, China
    3. Zhejiang Tianzheng Electric Co.,Ltd., Wenzhou 325604, China
    4. Zhejiang Juchuang Intelligent Technology Co.,Ltd., Wenzhou 325005, China
  • Received:2024-05-28 Online:2024-10-30 Published:2024-12-12

Abstract:

According to the breaking characteristics of DC arc,a method for predicting the electrical life of DC circuit breaker is proposed.According to the standard GB/T 14048.2—2020,the voltage waveform during the breaking process is obtained.After CEEMDAN filtering,the voltage signal within a period of arc break is used as the network training sample.The contact mass is used to represent the electrical life condition,and the contact mass decline curve is obtained by linear interpolation,which is used as the network training label.A one-dimensional convolutional structure prediction model of residual network(ResNet)is constructed using Pytorch framework.The experimental results show that the prediction error of the prosed method is less than 10%,which proves that the method can be used to predict the electrical life under GB/T 14048.2—2020 standard.

Key words: DC circuit breakers, electrical life prediction, residual network, one-dimensional convolution

CLC Number: