Scholarlink

<p>Condition-based monitoring and machine fault detection play vital roles in industry. However, it is challenging to identify subtle fault patterns as vibration signals are often affected by substantial noise interference. Autogram, as one of the effective diagnosis methods, relies on a single kurtosis indicator, making it noise-sensitive and suboptimal in selecting demodulation resonance bands. To address these limitations, the parameter-optimized multiscale permutation entropy index enhanced-gram (PMPEIgram) method is proposed by integrating the optimized multi-scale permutation entropy (MPE) with the kurtosis indicator, which offers improved robustness in demodulation resonance band selection. The genetic algorithm is employed to iteratively optimize MPE parameters, enhancing the detection ability in intense noise and weak faults. Validation with simulated and experimental datasets demonstrates the superiority of PMPEIgram in comparison with existing diagnostic methods, especially in identifying weak faults under challenging conditions.</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

PMPEIgram: An index-fused autogram approach for bearing diagnosis in high-noise and weak fault scenarios

  • 期刊论文

  • 发表日期：

    2026年2月9日

查看全文
  • Xiaoli Zhang,
  • Haopeng Han,
  • Yong Xiao,
  • Xin Luo

摘要

Condition-based monitoring and machine fault detection play vital roles in industry. However, it is challenging to identify subtle fault patterns as vibration signals are often affected by substantial noise interference. Autogram, as one of the effective diagnosis methods, relies on a single kurtosis indicator, making it noise-sensitive and suboptimal in selecting demodulation resonance bands. To address these limitations, the parameter-optimized multiscale permutation entropy index enhanced-gram (PMPEIgram) method is proposed by integrating the optimized multi-scale permutation entropy (MPE) with the kurtosis indicator, which offers improved robustness in demodulation resonance band selection. The genetic algorithm is employed to iteratively optimize MPE parameters, enhancing the detection ability in intense noise and weak faults. Validation with simulated and experimental datasets demonstrates the superiority of PMPEIgram in comparison with existing diagnostic methods, especially in identifying weak faults under challenging conditions.