Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

YOLO-based Bearing Fault Diagnosis With Continuous Wavelet Transform

Po-Heng Chou, Wei-Lung Mao, Ru-Ping Lin

TL;DR

The paper tackles bearing fault diagnosis from non-stationary vibration signals by converting 1D signals into 2D time-frequency spectrograms via Morlet-based continuous wavelet transform (CWT) and reframing diagnosis as object detection on the spectrograms. It introduces a three-stage CWT-YOLO pipeline: (1) generate CWT spectrograms, (2) prepare labeled spectrogram data with bounding boxes for four fault types, and (3) train YOLOv9–YOLOv11 to localize fault regions and classify faults. Across CWRU, PU, and IMS benchmarks, the approach achieves strong cross-dataset generalization and region-aware interpretability, with YOLOv11 offering the best practical balance of accuracy and efficiency (e.g., mAP@0.5 across datasets and notable PU performance due to the C2PSA attention). These results suggest that spectrogram-level detection provides a robust, interpretable alternative to traditional global classification for rotating machinery condition monitoring, enabling localized fault visualization and potential embedded deployment.

Abstract

This letter presents a locality-aware bearing fault diagnosis framework that operates on time-frequency representations and enables spatially interpretable decision-making. One-dimensional vibration signals are first mapped to two-dimensional time-frequency spectrograms using the continuous wavelet transform (CWT) with Morlet wavelets to enhance transient fault signatures. The diagnosis task is then formulated as object detection on the time-frequency plane, where YOLOv9, YOLOv10, and YOLOv11 are employed to localize fault-relevant regions and classify fault types simultaneously. Experiments on three public benchmarks, including Case Western Reserve University (CWRU), Paderborn University (PU), and Intelligent Maintenance System (IMS), demonstrate strong cross-dataset generalization compared with a representative MCNN-LSTM baseline. In particular, YOLOv11 achieves mAP@0.5 of 99.0% (CWRU), 97.8% (PU), and 99.5% (IMS), while providing region-aware visualization of fault patterns in the time-frequency domain. These results suggest that detection-based inference on CWT spectrograms provides an effective and interpretable complementary approach to conventional global classification for rotating machinery condition monitoring.

YOLO-based Bearing Fault Diagnosis With Continuous Wavelet Transform

TL;DR

The paper tackles bearing fault diagnosis from non-stationary vibration signals by converting 1D signals into 2D time-frequency spectrograms via Morlet-based continuous wavelet transform (CWT) and reframing diagnosis as object detection on the spectrograms. It introduces a three-stage CWT-YOLO pipeline: (1) generate CWT spectrograms, (2) prepare labeled spectrogram data with bounding boxes for four fault types, and (3) train YOLOv9–YOLOv11 to localize fault regions and classify faults. Across CWRU, PU, and IMS benchmarks, the approach achieves strong cross-dataset generalization and region-aware interpretability, with YOLOv11 offering the best practical balance of accuracy and efficiency (e.g., mAP@0.5 across datasets and notable PU performance due to the C2PSA attention). These results suggest that spectrogram-level detection provides a robust, interpretable alternative to traditional global classification for rotating machinery condition monitoring, enabling localized fault visualization and potential embedded deployment.

Abstract

This letter presents a locality-aware bearing fault diagnosis framework that operates on time-frequency representations and enables spatially interpretable decision-making. One-dimensional vibration signals are first mapped to two-dimensional time-frequency spectrograms using the continuous wavelet transform (CWT) with Morlet wavelets to enhance transient fault signatures. The diagnosis task is then formulated as object detection on the time-frequency plane, where YOLOv9, YOLOv10, and YOLOv11 are employed to localize fault-relevant regions and classify fault types simultaneously. Experiments on three public benchmarks, including Case Western Reserve University (CWRU), Paderborn University (PU), and Intelligent Maintenance System (IMS), demonstrate strong cross-dataset generalization compared with a representative MCNN-LSTM baseline. In particular, YOLOv11 achieves mAP@0.5 of 99.0% (CWRU), 97.8% (PU), and 99.5% (IMS), while providing region-aware visualization of fault patterns in the time-frequency domain. These results suggest that detection-based inference on CWT spectrograms provides an effective and interpretable complementary approach to conventional global classification for rotating machinery condition monitoring.

Paper Structure

This paper contains 14 sections, 1 equation, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. System Overview and Proposed CWT-YOLO
  3. Continuous Wavelet Transform (CWT)
  4. Time-Frequency Spectrogram Visualization
  5. Data Pre-processing and Labeling
  6. Signal Segmentation and Spectrogram Generation
  7. Bounding Box Annotation for Object Detection
  8. Dataset Splitting and Augmentation
  9. YOLO-based Fault Detection
  10. Experimental Results
  11. Performance Metrics
  12. Experimental Setup
  13. Performance Comparison Across Datasets
  14. Conclusion

Figures (2)

  • Figure 1: Sample vibration signals for four bearing conditions: (a) Normal, (b) Ball fault, (c) Inner race fault, (d) Outer race fault. Time (x-axis) and amplitude (y-axis) are shown for illustration, motivating a time-frequency transformation to expose discriminative fault signatures.
  • Figure 2: CWT-based spectrograms for four bearing conditions: (a) Normal, (b) Ball fault, (c) Inner race fault, (d) Outer race fault.