Scholarlink

<p>In real-world scenarios, railway bogie axle bearing diagnostic techniques face significant limitations in handling non-stationary operational conditions and multiple interfering noise sources, particularly in complex railway environments. Traditional approaches exhibit insufficient robustness and require complex post-processing strategies, especially when confronted with multivariate random pulse noise typical in railway operations. To address these challenges, this study proposes a novel Markovian spectral transition modeling framework with temporal dependencies, specifically designed for railway bogie axle bearing diagnostics in non-stationary transient environments. The framework introduces an innovative integration of Markovian modeling with multi-resolution wavelet analysis, alongside an amplitude-adaptive interference suppression mechanism that employs statistical signal modeling for dynamic thresholding. This comprehensive diagnostic methodology uniquely combines multiple signal processing techniques to handle transient interference noise in non-stationary bearing signals, leveraging intrinsic signal properties to enhance demodulation robustness. The proposed framework systematically integrates wavelet coefficients into Markovian state representations, establishing theoretical foundations for parameter optimization and providing a structured approach to railway bearing fault detection. The methodology demonstrates significant potential in advancing railway bogie axle bearing diagnostics, particularly in challenging operational environments characterized by complex noise patterns and non-stationary conditions typical of railway systems.</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

Markovian spectral transition modeling with temporal dependencies for railway bogie axle bearing diagnostics in non-stationary transient environments

  • 期刊论文

  • 发表日期：

    2026年2月24日

查看全文
  • Peng Chen,
  • Junxiao Ma,
  • Jia Gao,
  • Ge Xin,
  • Changbo He

摘要

In real-world scenarios, railway bogie axle bearing diagnostic techniques face significant limitations in handling non-stationary operational conditions and multiple interfering noise sources, particularly in complex railway environments. Traditional approaches exhibit insufficient robustness and require complex post-processing strategies, especially when confronted with multivariate random pulse noise typical in railway operations. To address these challenges, this study proposes a novel Markovian spectral transition modeling framework with temporal dependencies, specifically designed for railway bogie axle bearing diagnostics in non-stationary transient environments. The framework introduces an innovative integration of Markovian modeling with multi-resolution wavelet analysis, alongside an amplitude-adaptive interference suppression mechanism that employs statistical signal modeling for dynamic thresholding. This comprehensive diagnostic methodology uniquely combines multiple signal processing techniques to handle transient interference noise in non-stationary bearing signals, leveraging intrinsic signal properties to enhance demodulation robustness. The proposed framework systematically integrates wavelet coefficients into Markovian state representations, establishing theoretical foundations for parameter optimization and providing a structured approach to railway bearing fault detection. The methodology demonstrates significant potential in advancing railway bogie axle bearing diagnostics, particularly in challenging operational environments characterized by complex noise patterns and non-stationary conditions typical of railway systems.