Scholarlink

<p>The vibration signal of a rolling bearing fault is affected by transmission path effects, Gaussian noise, random impacts, and other interferences during propagation, which mask the transient features and modulated frequency components, making it difficult to identify and fault feature extraction. When using feature mode decomposition (FMD), a priori parameter settings have a significant impact on the decomposition results, as improper settings may lead to issues such as mode mixing and the loss of key fault features. Meanwhile, existing noise reduction algorithm exhibits poor performance in terms of robustness to strong noise and retention of complete fault information. To address these challenges, this study proposes a novel approach for extracting fault features, which leverages the improved hippo optimization (IHO) algorithm to optimize the parameters of FMD, that is IHO-FMD. The IHO algorithm integrates logistic chaotic mapping, opposition-based learning, and adaptive weighting strategy, and uses envelope entropy as the objective function to determine the optimal decomposition parameters in FMD, thereby enabling effective decomposition of the original signal. Subsequently, kurtosis is employed to select the most informative intrinsic mode function (IMF) for envelope demodulation. Experimental studies on the CWRU and XJTU-SY datasets show that the method proposed in this study has stronger parameter adaptability and fault diagnosis accuracy under noisy conditions. Compared with existing EMD, VMD, and traditional FMD algorithm, this method performs better in extracting fault feature frequencies and can effectively identify bearing faults.</p>

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

A novel bearing fault feature extraction algorithm based on improved hippo optimization feature mode decomposition

  • 期刊论文

  • 发表日期：

    2025年7月11日

查看全文
  • Shoubin Wang,
  • Zhilong Yu,
  • Guili Peng,
  • Shaojie Yang,
  • Huanyu Wu,
  • Youbing Li,
  • Lewei Jing,
  • Xinchang Fang

摘要

The vibration signal of a rolling bearing fault is affected by transmission path effects, Gaussian noise, random impacts, and other interferences during propagation, which mask the transient features and modulated frequency components, making it difficult to identify and fault feature extraction. When using feature mode decomposition (FMD), a priori parameter settings have a significant impact on the decomposition results, as improper settings may lead to issues such as mode mixing and the loss of key fault features. Meanwhile, existing noise reduction algorithm exhibits poor performance in terms of robustness to strong noise and retention of complete fault information. To address these challenges, this study proposes a novel approach for extracting fault features, which leverages the improved hippo optimization (IHO) algorithm to optimize the parameters of FMD, that is IHO-FMD. The IHO algorithm integrates logistic chaotic mapping, opposition-based learning, and adaptive weighting strategy, and uses envelope entropy as the objective function to determine the optimal decomposition parameters in FMD, thereby enabling effective decomposition of the original signal. Subsequently, kurtosis is employed to select the most informative intrinsic mode function (IMF) for envelope demodulation. Experimental studies on the CWRU and XJTU-SY datasets show that the method proposed in this study has stronger parameter adaptability and fault diagnosis accuracy under noisy conditions. Compared with existing EMD, VMD, and traditional FMD algorithm, this method performs better in extracting fault feature frequencies and can effectively identify bearing faults.