Scholarlink

To tackle the complexity of motor bearing fault signals and the challenges in accurately extracting fault features, a fault feature extraction method for motor bearings is proposed based on an enhanced blood-sucking leech optimization (EBSLO) algorithm to optimize the parameters of variational mode decomposition (VMD). Firstly, an improved EBSLO algorithm is introduced to strengthen the adaptive optimization performance of the conventional blood-sucking leech optimization (BSLO) method. Secondly, the EBSLO algorithm is used to obtain the optimal parameter combination VMD instead of the traditional empirical parameter setting VMD to decompose the original fault signal by intrinsic mode function (IMF) and replace the empirical parameter setting of VMD. Finally, IMF components with a kurtosis value of at least 3 are selected for reconstruction, and the reconstructed signal is analyzed using Hilbert envelope spectrum analysis. The simulation results show that the proposed method successfully reduces IMF mode aliasing resulting from incorrect VMD parameter settings, and can accurately extract the fault characteristics of motor bearing from the original fault signal, supporting precise fault diagnosis.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Fault Feature Extraction of Motor Bearing Based on EBSLO-VMD

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Guihua Zou,
  • Yuan Xie,
  • Wenxian Yang

摘要

To tackle the complexity of motor bearing fault signals and the challenges in accurately extracting fault features, a fault feature extraction method for motor bearings is proposed based on an enhanced blood-sucking leech optimization (EBSLO) algorithm to optimize the parameters of variational mode decomposition (VMD). Firstly, an improved EBSLO algorithm is introduced to strengthen the adaptive optimization performance of the conventional blood-sucking leech optimization (BSLO) method. Secondly, the EBSLO algorithm is used to obtain the optimal parameter combination VMD instead of the traditional empirical parameter setting VMD to decompose the original fault signal by intrinsic mode function (IMF) and replace the empirical parameter setting of VMD. Finally, IMF components with a kurtosis value of at least 3 are selected for reconstruction, and the reconstructed signal is analyzed using Hilbert envelope spectrum analysis. The simulation results show that the proposed method successfully reduces IMF mode aliasing resulting from incorrect VMD parameter settings, and can accurately extract the fault characteristics of motor bearing from the original fault signal, supporting precise fault diagnosis.