416 / 2022-03-15 21:02:52
A Mechanical Vibration Fault Diagnosis Method of Generator Outlet Circuit Breaker Based on Deep Residual Shrinkage Network
Geerator outlet circuit breaker,mechanical vibration fault,deep residual shrinkage network
摘要录用
Guowen Hao / State Grid Xinyuan Holdings Co., Ltd.
Erhan Hu / Chongqing University
Yaxiong Tan / Chongqing University
The operation frequency of generator outlet circuit breaker (GOCB) in pumped storage power stations are approximately a thousand times higher than that in conventional hydropower stations. As we know, the mechanical failure accounts for a high proportion of all the faults. Therefore, it is of great practical significance to diagnose the mechanical characteristics of GOCB. However, due to high operation frequency, the vibration signal is much more complex and nonlinear, which makes it difficult to diagnose the mechanical fault. This paper proposes a mechanical fault diagnosis method of GOCB based on deep residual shrinkage network. Firstly, the original vibration signal is segmented in time domain. Each part after segmentation is integrated to reflect the characteristics of vibration signal amplitude, phase and frequency change. Four typical fault types are simulated in the laboratory, and the vibration characteristic database under normal and four fault conditions is established. Then in the training model, a sub-network is applied to automatically set the threshold in the soft threshold, which can enhance the efficiency of feature training and improve the accuracy of fault diagnosis. Finally, the fault diagnosis method is verified by test data. The results show that The diagnostic accuracy of this method reaches 97%. Experiments on a 24 kV GOCB verify the effectiveness of the proposed method.
重要日期
  • 会议日期

    09月25日

    2022

    09月29日

    2022

  • 08月15日 2022

    提前注册日期

  • 09月10日 2022

    报告提交截止日期

  • 11月10日 2022

    注册截止日期

  • 11月30日 2022

    初稿截稿日期

  • 11月30日 2022

    终稿截稿日期

主办单位
IEEE DEIS
承办单位
Chongqing University
移动端
在手机上打开
小程序
打开微信小程序
客服
扫码或点此咨询