EUVIMEDEUVIMED entdecken
Crossref

Remaining Useful Life Prediction for ATS Bearings Based on a Temporal Convolutional Network Fused With Autoregression

Runxia Guo, Xianfeng Luo, Siying Wang, Chao Huang

Quality and Reliability Engineering International · 2026-04 · 42 (3) · 1321-1332

DOI: 10.1002/qre.70147 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Quality and Reliability Engineering International, online erschienen am 28.12.2025. Verfasst von Runxia Guo und 3 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 28 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The air turbine starter (ATS) initiates an aircraft's engine before takeoff, with rolling bearings being crucial components. Predicting the remaining useful life (RUL) of these bearings is challenging due to their complex working conditions. This study presents an RUL prediction methodology that combines deep learning networks and autoregression (AR). The deep learning network capitalizes on its high prediction accuracy for short‐term prediction, while AR leverages its robustness and resistance to interference to depict the long‐term declining trend in RUL. The fusion of these two approaches enables precise RUL prediction. The network model introduced in this paper integrates a temporal convolutional network (TCN) with a lightweight channel attention mechanism (LCAM) and runs them in parallel with a nonlinear neural network AR module. The outputs from both models are summed up to generate final prediction results, establishing an end‐to‐end RUL prediction framework. The efficacy of this methodology is confirmed through its application to the laboratory mechanical components accelerated aging fatigue test rig dataset. A comparative analysis reveals that the proposed AR‐LCAM‐TCN framework surpasses other state‐of‐the‐art methods in terms of RUL and system prediction accuracy, as well as computational efficiency.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Guo, R., Luo, X., Wang, S., & Huang, C. (2026). Remaining Useful Life Prediction for ATS Bearings Based on a Temporal Convolutional Network Fused With Autoregression. Quality and Reliability Engineering International, 42(3), 1321-1332. https://doi.org/10.1002/qre.70147

Vancouver

Guo R, Luo X, Wang S, Huang C. Remaining Useful Life Prediction for ATS Bearings Based on a Temporal Convolutional Network Fused With Autoregression. Quality and Reliability Engineering International. 2026;42(3):1321-1332. doi:10.1002/qre.70147

BibTeX
@article{Guo2026,
  author = {Guo, Runxia and Luo, Xianfeng and Wang, Siying and Huang, Chao},
  title = {Remaining Useful Life Prediction for ATS Bearings Based on a Temporal Convolutional Network Fused With Autoregression},
  journal = {Quality and Reliability Engineering International},
  year = {2026},
  volume = {42},
  number = {3},
  pages = {1321--1332},
  doi = {10.1002/qre.70147}
}
Kostenlos starten →