Prior knowledge-driven meta-learning for fault detection under varying working conditions
ISA Transactions · 2026-10
DOI: 10.1016/j.isatra.2026.10.025 ↗Kurz gefasst
Fachartikel in der Zeitschrift ISA Transactions, 2026-10. Verfasst von Yuan Xu und 9 weiteren Autor:innen. Beteiligte Einrichtung: College of Information Science & Technology, Beijing University of Chemical Technology, Beijing, 100029, China; Engineering Research Center of Intell…. Die Arbeit nennt 40 Literaturangaben. Gefördert von: Beijing Natural Science Foundation, National Natural Science Foundation of China.
Abstract (Originalsprache)
This paper proposes a novel prior knowledge-driven meta-learning (PKML) fault detection approach to address practical scenarios involving scarce fault samples under varying working conditions. Specifically, the PKML method is built upon a model-agnostic meta-learning framework, where meta-task samples from diverse working conditions are mapped into embedding vectors to obtain a unified representation of multi-condition features. To enhance the discriminability of the embedding representations, a learnable sample-attention mechanism is incorporated to adjust the relative contributions of support samples during condition embedding construction. Furthermore, a wide-and-deep convolutional neural network is employed as the base classifier for meta-learning, with embedded conditional layers to enhance adaptability to varying working conditions. Through a feature-wise linear modulation mechanism, the classifier parameters are dynamically adjusted according to condition-aware structural information, enabling effective adaptation and generalization to unseen working conditions. Finally, experimental results on three representative bearing fault datasets demonstrate that the proposed PKML method consistently outperforms existing approaches under varying working conditions and few-shot scenarios.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- College of Information Science & Technology, Beijing University of Chemical Technology, Beijing, 100029, China; Engineering Research Center of Intelligent PSE, Ministry of Education of China, Beijing, 100029, China. Ele…
- CHN Energy Technology & Environment Limited, Beijing, 100039, China; CHN Energy ZhiShen Control Technology Co., Ltd., Beijing, 102211, China; State R & D Center of Control System and Information Security Technologies fo…
- Research Institute of Mine Artificial Intelligence, Chinese Institute of Coal Science, Beijing, 100013, China. Electronic address: luoyi@ccteg-bigdata.com.
- Faculty of Applied Science, Macao Polytechnic University, Macao SAR, 999078, China. Electronic address: wke@mpu.edu.mo.
Förderung
- Beijing Natural Science Foundation
- National Natural Science Foundation of China
Zitieren
APA
Xu, Y., Wu, K. Y., Liu, M. J., Ma, F., Luo, Y., Ke, W., He, Y. L., Zhu, Q. X., Zhang, Y., & Zhang, M. Q. (2026). Prior knowledge-driven meta-learning for fault detection under varying working conditions. ISA Transactions. https://doi.org/10.1016/j.isatra.2026.10.025
Vancouver
Xu Y, Wu KY, Liu MJ, Ma F, Luo Y, Ke W, et al. Prior knowledge-driven meta-learning for fault detection under varying working conditions. ISA Transactions. 2026. doi:10.1016/j.isatra.2026.10.025
BibTeX
@article{Xu2026,
author = {Xu, Yuan and Wu, Ke-Yu and Liu, Ming-Jia and Ma, Feng and Luo, Yi and Ke, Wei and He, Yan-Lin and Zhu, Qun-Xiong and Zhang, Yang and Zhang, Ming-Qing},
title = {Prior knowledge-driven meta-learning for fault detection under varying working conditions},
journal = {ISA Transactions},
year = {2026},
doi = {10.1016/j.isatra.2026.10.025}
}
EUVIMED