EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

WCFormer: A wavelet-enhanced CNN-transformer hybrid network for bearing fault diagnosis using multi-sensor signal fusion

Xinjie Cao, Zenggui Gao, Hongjiang Lu, Jingwei Tang, Yuyan Yao, Lilan Liu

Journal of Computational Design and Engineering · 02.09.2025 · 12 (9) · 82-100

DOI: 10.1093/jcde/qwaf080 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Computational Design and Engineering, online erschienen am 12.09.2025. Verfasst von Xinjie Cao und 5 weiteren Autor:innen. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Digital Twin Acoustic Fault Diagnosis and Control Technology and Application for Deep-sea Ships.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Bearings are core components of rotating machinery, and their failures can cause significant production accidents. Current multi-source data fusion methods and independent network architectures show limited diagnostic performance on small and medium-sized datasets. To address multi-sensor data fusion and feature extraction in bearing fault diagnosis, we propose MCWT-WCFormer, a lightweight convolution-Transformer hybrid network with three key modules: multi-channel wavelet transform (MCWT), channel-spatial attention net (CSAN), and wave-convolution and attention fusion net (WAFN). (1) MCWT transforms multi-source signals into grayscale images through wavelet transform, stacks them into RGB format, integrating time-frequency information. (2) CSAN generates spatial and channel information descriptors and dynamically weights feature maps. (3) WAFN extracts high-frequency features of the Transformer’s Key by introducing wavelet transform convolution, realizing joint learning of local-global features. MCWT-WCFormer optimizes efficiency and performance by leveraging the inductive bias of convolutional neural network and the scalability of the Transformer. Cross-evaluated on HUST-gearbox (Huazhong University of Science and Technology gearbox) and SHU-TSTB (Shanghai University Transmission Simulation Test Bench) datasets, MCWT-WCFormer achieves 98.12% ± 0.17% and 98.03% ± 0.12% accuracy, respectively, with a single sample diagnosis time of about 4.2 ms while having the lowest complexity (43.25 GFLOPs) and parameters (5.91 M). It can be integrated into industrial digital twin systems cost-effectively while supporting new metrics like cyclic-correntropy. It is extendable to rotating machinery health management like steam turbines and wind turbine gearboxes.

Förderung

  • Digital Twin Acoustic Fault Diagnosis and Control Technology and Application for Deep-sea Ships

Zitieren

APA

Cao, X., Gao, Z., Lu, H., Tang, J., Yao, Y., & Liu, L. (2025). WCFormer: A wavelet-enhanced CNN-transformer hybrid network for bearing fault diagnosis using multi-sensor signal fusion. Journal of Computational Design and Engineering, 12(9), 82-100. https://doi.org/10.1093/jcde/qwaf080

Vancouver

Cao X, Gao Z, Lu H, Tang J, Yao Y, Liu L. WCFormer: A wavelet-enhanced CNN-transformer hybrid network for bearing fault diagnosis using multi-sensor signal fusion. Journal of Computational Design and Engineering. 2025;12(9):82-100. doi:10.1093/jcde/qwaf080

BibTeX
@article{Cao2025,
  author = {Cao, Xinjie and Gao, Zenggui and Lu, Hongjiang and Tang, Jingwei and Yao, Yuyan and Liu, Lilan},
  title = {WCFormer: A wavelet-enhanced CNN-transformer hybrid network for bearing fault diagnosis using multi-sensor signal fusion},
  journal = {Journal of Computational Design and Engineering},
  year = {2025},
  volume = {12},
  number = {9},
  pages = {82--100},
  doi = {10.1093/jcde/qwaf080}
}
Kostenlos starten →