Physics-Topology-Anchored Learning: A Robust and Lightweight Framework for Time-Series Prediction and Anomaly Detection Under Data Scarcity
Sensors · 09.03.2026 · 26 (5) · 1721
DOI: 10.3390/s26051721 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 09.03.2026. Verfasst von Xuanhao Hua und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Future Technology, Xi'an Jiaotong University, Xi'an 710049, China. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Hangzhou Special Equipment Inspection and Research Institute, Key Technology Research on Risk Monitoring, Intelligent Diagnosis, and Maintenance Decision-making for Escalators Based on Multimodal Large Models.
Abstract (Originalsprache)
Health monitoring of complex systems is critical for ensuring reliability and achieving cost-effective reusability. However, deploying deep learning models in this domain is impeded by two primary constraints: the scarcity of high-quality fault samples and the restricted computational resources available on-board. To address these challenges, this paper proposes a Physics-Topology-Anchored Learning (PTAL) framework. The core innovation lies in the effective integration of physical inductive bias into the model architecture. Specifically, PTAL incorporates a predefined adjacency matrix, derived from the physical mechanism, as a structural prior. This design anchors the neural network to explicit physical causality, effectively constraining the hypothesis space and reducing the model’s dependency on large-scale data. Furthermore, by coupling this physics-informed structure with a lightweight recurrent attention mechanism, the model avoids the high computational overhead typical of generic large-scale networks. Experimental evaluations demonstrate that PTAL achieves a peak diagnostic accuracy of 97.8% and a low standard deviation of 0.1145, significantly outperforming baseline models in data-scarce regimes. The results confirm that the proposed model successfully leverages physical bias to maintain a favorable trade-off between diagnostic performance and computational efficiency, making it highly suitable for the resource-constrained environments of complex systems.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Future Technology, Xi'an Jiaotong University, Xi'an 710049, China.
- Hangzhou Special Equipment Inspection and Science Research Institute, Hangzhou 310051, China.
- School of Mechanical Engineering, Xi'an Jiaotong University, Xi'an 710049, China.
Förderung
- Hangzhou Special Equipment Inspection and Research Institute
- Key Technology Research on Risk Monitoring, Intelligent Diagnosis, and Maintenance Decision-making for Escalators Based on Multimodal Large Models
Zitieren
APA
Hua, X., Yin, W., Wang, L., Ma, M., Yuan, J., & Zhang, J. (2026). Physics-Topology-Anchored Learning: A Robust and Lightweight Framework for Time-Series Prediction and Anomaly Detection Under Data Scarcity. Sensors, 26(5), 1721. https://doi.org/10.3390/s26051721
Vancouver
Hua X, Yin W, Wang L, Ma M, Yuan J, Zhang J. Physics-Topology-Anchored Learning: A Robust and Lightweight Framework for Time-Series Prediction and Anomaly Detection Under Data Scarcity. Sensors. 2026;26(5):1721. doi:10.3390/s26051721
BibTeX
@article{Hua2026,
author = {Hua, Xuanhao and Yin, Weiqi and Wang, Libin and Ma, Meng and Yuan, Jianfeng and Zhang, Jing},
title = {Physics-Topology-Anchored Learning: A Robust and Lightweight Framework for Time-Series Prediction and Anomaly Detection Under Data Scarcity},
journal = {Sensors},
year = {2026},
volume = {26},
number = {5},
pages = {1721},
doi = {10.3390/s26051721}
}
EUVIMED