EUVIMEDEUVIMED entdecken
Crossref

Global-local feature extraction network incorporating data fluctuation augmentation for equipment remaining useful life prediction

Ziyang Zheng, Changhua Hu, Zhaoqiang Wang, Tao Peng, Mingzhe Leng, Chuanyang Li, Zeming Zhang

Measurement Science and Technology · 08.05.2026 · 37 (19) · 195106

DOI: 10.1088/1361-6501/ae65bd ↗

Kurz gefasst

Fachartikel in der Zeitschrift Measurement Science and Technology, online erschienen am 12.05.2026. Verfasst von Ziyang Zheng und 6 weiteren Autor:innen. Die Arbeit nennt 53 Literaturangaben. Gefördert von: Open Foundation of the State Key Laboratory of Fluid Power and Mechatronic Systems, School-level Scientific Research Project of China Fire and Rescue Academy, National Key Research and Development Program of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Remaining useful life (RUL) prediction is recognized as the key technique for system predictive maintenance which can effectively mitigate unexpected failures and reduce operational costs. This paper proposes a novel global-local feature extraction network incorporating data fluctuation feature augmentation (abbreviated as FEFA here) for equipment RUL prediction. Specifically, the temporal convolutional networks and Transformer encoders are collaboratively utilized to extract the overarching long-term degradation trends, while a novel causal convolutional gated unit is designed to capture the dynamic short-term fluctuation features. Additionally, a 2D squeeze-and-excitation fusion block, inspired by channel attention mechanisms, is employed to effectively integrate these two distinct dynamics, thereby significantly improving the RUL prediction accuracy. The proposed FEFA model has been thoroughly evaluated on two equipment datasets. The results demonstrate that the FEFA model exhibits powerful feature extraction capabilities, offering a robust and efficient solution for accurate RUL prediction.

Förderung

  • Open Foundation of the State Key Laboratory of Fluid Power and Mechatronic Systems
  • School-level Scientific Research Project of China Fire and Rescue Academy
  • National Key Research and Development Program of China
  • Key Project of the Shaanxi Province Natural Science Foundation
  • National Natural Science Foundation of China

Zitieren

APA

Zheng, Z., Hu, C., Wang, Z., Peng, T., Leng, M., Li, C., & Zhang, Z. (2026). Global-local feature extraction network incorporating data fluctuation augmentation for equipment remaining useful life prediction. Measurement Science and Technology, 37(19), 195106. https://doi.org/10.1088/1361-6501/ae65bd

Vancouver

Zheng Z, Hu C, Wang Z, Peng T, Leng M, Li C, et al. Global-local feature extraction network incorporating data fluctuation augmentation for equipment remaining useful life prediction. Measurement Science and Technology. 2026;37(19):195106. doi:10.1088/1361-6501/ae65bd

BibTeX
@article{Zheng2026,
  author = {Zheng, Ziyang and Hu, Changhua and Wang, Zhaoqiang and Peng, Tao and Leng, Mingzhe and Li, Chuanyang and Zhang, Zeming},
  title = {Global-local feature extraction network incorporating data fluctuation augmentation for equipment remaining useful life prediction},
  journal = {Measurement Science and Technology},
  year = {2026},
  volume = {37},
  number = {19},
  pages = {195106},
  doi = {10.1088/1361-6501/ae65bd}
}
Kostenlos starten →