EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Fault diagnosis of wind turbine gearboxes based on optimized VMD with transformer-KAN

Yugang Li, Pingyuan Ma, Hongtao Sun, Zheng Chen

Engineering Research Express · 01.06.2026 · 8 (11) · 115525

DOI: 10.1088/2631-8695/ae76bb ↗

Kurz gefasst

Fachartikel in der Zeitschrift Engineering Research Express, online erschienen am 11.06.2026. Verfasst von Yugang Li und 3 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Rizhao Natural Science Foundation, National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The diagnosis of wind turbine gearbox faults using noise signals is an effective method. However, the gearbox is in high-speed operation for a long time, which generates other noises and mixes the fault signal with the noise, which is difficult to detect. To address this issue, this paper presents a feature extraction method that leverages the rime optimization algorithm (RIME) to enhance the performance of variational mode decomposition (VMD); and combines with transformer-KAN (Kolmogorov–Arnold Network, KAN) to construct a diagnostic model, which is applied to thein wind turbine gearbox fault diagnosis. The KAN, inspired by the Kolmogorov–Arnold Representation Theorem, is a novel neural network architecture capable of learning complex nonlinear mapping relationships with enhanced interpretability. Firstly, RIME is used to optimize the signal of the gearbox data set, and permutation entropy is chosen as the fitness function, the fault signal undergoes VMD after determining the optimal number of modes K and the penalty factor α through computational optimization, and finally, the decomposed signals are fed into the Transformer-KAN diagnostic model for the diagnosis of wind turbine gearbox faults. The experimental findings reveal that the computationally optimized parameters of this method deliver the most effective noise suppression performance, and the fault diagnosis accuracy can reach 99%, which proves the feasibility of the proposed wind turbine gearbox fault diagnosis model.

Förderung

  • Rizhao Natural Science Foundation
  • National Natural Science Foundation of China

Zitieren

APA

Li, Y., Ma, P., Sun, H., & Chen, Z. (2026). Fault diagnosis of wind turbine gearboxes based on optimized VMD with transformer-KAN. Engineering Research Express, 8(11), 115525. https://doi.org/10.1088/2631-8695/ae76bb

Vancouver

Li Y, Ma P, Sun H, Chen Z. Fault diagnosis of wind turbine gearboxes based on optimized VMD with transformer-KAN. Engineering Research Express. 2026;8(11):115525. doi:10.1088/2631-8695/ae76bb

BibTeX
@article{Li2026,
  author = {Li, Yugang and Ma, Pingyuan and Sun, Hongtao and Chen, Zheng},
  title = {Fault diagnosis of wind turbine gearboxes based on optimized VMD with transformer-KAN},
  journal = {Engineering Research Express},
  year = {2026},
  volume = {8},
  number = {11},
  pages = {115525},
  doi = {10.1088/2631-8695/ae76bb}
}
Kostenlos starten →