EUVIMEDEUVIMED entdecken
Crossref

Time–frequency convolutional-Transformer network with attention fusion for rotor system fault diagnosis

Dahai Zhang, Youming Hao, Hongyu Li, Cunzhong Gong, Junzeng He, Qingguo Fei

Measurement Science and Technology · 27.03.2026 · 37 (12) · 126101

DOI: 10.1088/1361-6501/ae4f0d ↗

Kurz gefasst

Fachartikel in der Zeitschrift Measurement Science and Technology, online erschienen am 26.03.2026. Verfasst von Dahai Zhang und 5 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 40 Literaturangaben. Gefördert von: Jiangsu Natural Science Fund, National Natural Science Foundation of China, Jiangsu Funding Program.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Under strong noise interference and complex operating conditions, conventional signal processing, machine learning, and deep learning approaches often struggle to effectively capture the multi-scale and nonlinear characteristics of rotor vibration signals. To address these challenges, this paper proposes a time–frequency convolutional-Transformer network with attention fusion (TFCTA) for intelligent rotor system fault diagnosis. TFCTA integrates a dual-path multi-scale time-domain extractor, a frequency-domain convolutional-Transformer encoder, and a cross-domain attention fusion mechanism, to jointly model local patterns, global time–frequency dependencies, and inter-domain interactions. These components collectively enable effective local-pattern learning, global time–frequency modelling and adaptive cross-domain feature fusion. The fused representation is finally processed by a main-auxiliary joint classifier to improve convergence stability and enhance discriminative capability. Experiments on both a turbine-rotor simulation dataset and a publicly available rotor database from Wuhan University demonstrate that TFCTA maintains high robustness and strong generalisation capability across a wide range of SNR conditions, highlighting its effectiveness under complex and noisy operating environments.

Förderung

  • Jiangsu Natural Science Fund
  • National Natural Science Foundation of China
  • Jiangsu Funding Program

Zitieren

APA

Zhang, D., Hao, Y., Li, H., Gong, C., He, J., & Fei, Q. (2026). Time–frequency convolutional-Transformer network with attention fusion for rotor system fault diagnosis. Measurement Science and Technology, 37(12), 126101. https://doi.org/10.1088/1361-6501/ae4f0d

Vancouver

Zhang D, Hao Y, Li H, Gong C, He J, Fei Q. Time–frequency convolutional-Transformer network with attention fusion for rotor system fault diagnosis. Measurement Science and Technology. 2026;37(12):126101. doi:10.1088/1361-6501/ae4f0d

BibTeX
@article{Zhang2026,
  author = {Zhang, Dahai and Hao, Youming and Li, Hongyu and Gong, Cunzhong and He, Junzeng and Fei, Qingguo},
  title = {Time–frequency convolutional-Transformer network with attention fusion for rotor system fault diagnosis},
  journal = {Measurement Science and Technology},
  year = {2026},
  volume = {37},
  number = {12},
  pages = {126101},
  doi = {10.1088/1361-6501/ae4f0d}
}
Kostenlos starten →