Dynamic pseudo-label guided adversarial multi-scale graph convolutional network for cross-domain fault diagnosis
Review of Scientific Instruments · 01.08.2026 · 97 (8)
DOI: 10.1063/5.0331532 ↗Kurz gefasst
Fachartikel in der Zeitschrift Review of Scientific Instruments, online erschienen am 13.08.2026. Verfasst von Jinqi Gao und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Artificial Intelligence, Guilin University of Aerospace Technology, Guilin 541004, China. Die Arbeit nennt 45 Literaturangaben. Gefördert von: Guangxi Natural Science Foundation, National Natural Science Foundation of China, Key Laboratory of the State Administration for Market Regulation.
Abstract (Originalsprache)
Recently, domain adaptation, a form of transfer learning, has been extensively applied to mechanical fault diagnosis across diverse operating conditions to address the challenges of insufficient labeled data and frequent operational state transitions. However, most existing approaches focus solely on the spatial distribution of inter-domain categorical features, neglecting the clustering properties of feature clusters. Moreover, most pseudo-labeling approaches lack a rational screening mechanism based on the quality of the pseudo-labels. To address these challenges, this paper proposes a Dynamic Pseudo-Label Guided Adversarial Multi-Scale Graph Convolutional Network (DPAMGCN) for unsupervised cross-domain fault diagnosis. First, a network architecture is designed by cascading a multi-scale parallel Convolutional Neural Network (CNN) with a multi-receptive-field graph convolutional network (MRF-GCN) to extract features. Second, a multi-objective total-loss function is constructed that integrates geometric loss with three standard loss functions to jointly optimize feature clustering and spatial distribution. Finally, a dynamic threshold-based pseudo-label filtering strategy is proposed that, when combined with a geometric loss, enhances the model’s generalization capability. Cross-domain transfer experiments conducted on the University of Ottawa (Ottawa) and Huazhong University of Science and Technology (HUST) benchmark datasets demonstrate that DPAMGCN achieves outstanding cross-domain diagnostic performance under the proposed optimization strategy and pseudo-label screening mechanism.
Autor:innen und Institutionen
- School of Artificial Intelligence, Guilin University of Aerospace Technology, Guilin 541004, China.
- Guangxi Key Laboratory of Special Engineering Equipment and Control, Guilin 541004, China.
- School of Mechanical and Electrical Engineering, Lanzhou University of Technology, Lanzhou 730050, China.
- School of Mechanical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China.
- School of Automation, Wuhan University of Technology, Wuhan 430070, China.
ORCID-Profile: Jinqi Gao, Bo Zhang, Tianlong Huo, Xiaoli Zhao, Tao Liu
Förderung
- Guangxi Natural Science Foundation
- National Natural Science Foundation of China
- Key Laboratory of the State Administration for Market Regulation
- Open Fund of Intelligent Control Laboratory
- Open Fund of Laboratory of Science and Technology on Integrated Logistics Support, National University of Defense Technology
Zitieren
APA
Gao, J., Zhang, B., Huo, T., Zhao, X., Du, H., & Liu, T. (2026). Dynamic pseudo-label guided adversarial multi-scale graph convolutional network for cross-domain fault diagnosis. Review of Scientific Instruments, 97(8). https://doi.org/10.1063/5.0331532
Vancouver
Gao J, Zhang B, Huo T, Zhao X, Du H, Liu T. Dynamic pseudo-label guided adversarial multi-scale graph convolutional network for cross-domain fault diagnosis. Review of Scientific Instruments. 2026;97(8). doi:10.1063/5.0331532
BibTeX
@article{Gao2026,
author = {Gao, Jinqi and Zhang, Bo and Huo, Tianlong and Zhao, Xiaoli and Du, Hongtai and Liu, Tao},
title = {Dynamic pseudo-label guided adversarial multi-scale graph convolutional network for cross-domain fault diagnosis},
journal = {Review of Scientific Instruments},
year = {2026},
volume = {97},
number = {8},
doi = {10.1063/5.0331532}
}
EUVIMED