Scholarlink

Abstract <p>Diagnosing incipient faults in rotating machinery is crucial for reducing maintenance costs and preventing the occurrence of other faults. Accurately extracting fault-related features from vibrational time series is fundamental to early fault detection. Entropy has emerged as a robust metric for quantifying complexity in time series; however, entropy-based feature extraction methods often require parameter tuning and tend to overlook frequency-domain features. This paper introduces BiEntropiGram, a novel time-frequency entropy mapping technique synthesizing information from both the time and frequency domains into a 2D image representation. In this framework, time-domain features are extracted using time shift multi-scale attention entropy (TSMAtE), while frequency-domain features are derived through time shift multi-scale spectral entropy (TSMSpE). Noise analysis determines the optimal number of time scales and the ideal time series length to optimize performance. Analytical studies with various noise types further demonstrate the superior performance of BiEntropiGram compared to features extracted solely via TSMAtE or TSMSpE. The proposed fault diagnosis system utilizes an optimized convolutional neural network (CNN) coupled with the Bees Algorithm (BA)—termed BACNN—to refine feature selection and classify faults using an optimized support vector machine. The efficacy of BiEntropiGram is validated across five publicly available datasets, achieving 100% diagnostic accuracy on three datasets with faults of varying severity. Furthermore, BiEntropiGram outperforms traditional methods, including the Morlet and Mexican hat continuous wavelet transforms, short-time Fourier transform, and Wigner–Ville distribution. The performance of the BiEntropiGram in incipient fault diagnosis is also compared with the state-of-the-art methods. These results confirm the effectiveness of the proposed BiEntropiGram-based fault diagnosis approach for diagnosing incipient faults in rotating machinery.</p> Graphical abstract <p></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

BiEntropiGram: a novel time-frequency entropy map for rotating machines incipient fault diagnosis

  • 期刊论文

  • 发表日期：

    2025年5月6日

查看全文
  • Emadaldin Sh Khoram-Nejad,
  • Abdolreza Ohadi,
  • Farshad Almasganj

摘要

Abstract

Diagnosing incipient faults in rotating machinery is crucial for reducing maintenance costs and preventing the occurrence of other faults. Accurately extracting fault-related features from vibrational time series is fundamental to early fault detection. Entropy has emerged as a robust metric for quantifying complexity in time series; however, entropy-based feature extraction methods often require parameter tuning and tend to overlook frequency-domain features. This paper introduces BiEntropiGram, a novel time-frequency entropy mapping technique synthesizing information from both the time and frequency domains into a 2D image representation. In this framework, time-domain features are extracted using time shift multi-scale attention entropy (TSMAtE), while frequency-domain features are derived through time shift multi-scale spectral entropy (TSMSpE). Noise analysis determines the optimal number of time scales and the ideal time series length to optimize performance. Analytical studies with various noise types further demonstrate the superior performance of BiEntropiGram compared to features extracted solely via TSMAtE or TSMSpE. The proposed fault diagnosis system utilizes an optimized convolutional neural network (CNN) coupled with the Bees Algorithm (BA)—termed BACNN—to refine feature selection and classify faults using an optimized support vector machine. The efficacy of BiEntropiGram is validated across five publicly available datasets, achieving 100% diagnostic accuracy on three datasets with faults of varying severity. Furthermore, BiEntropiGram outperforms traditional methods, including the Morlet and Mexican hat continuous wavelet transforms, short-time Fourier transform, and Wigner–Ville distribution. The performance of the BiEntropiGram in incipient fault diagnosis is also compared with the state-of-the-art methods. These results confirm the effectiveness of the proposed BiEntropiGram-based fault diagnosis approach for diagnosing incipient faults in rotating machinery.

Graphical abstract