Treffer: Global receptive field graph attention network for unsupervised domain adaptation fault diagnosis in variable operating conditions.

While deep learning has advanced significantly in machinery diagnosis, models trained on source domain data struggle with real-world applications due to varying operating conditions in the target domain. To address this, we propose a novel solution, the Global Receptive Field-based Graph Attention Network (GRF-GAT), for the fault diagnosis of varying conditions by the scheme of unsupervised domain adaptation. Unlike existing methods, GRF-GAT models class labels, domain labels, and associations and distributions among samples within a unified deep network. GRF-GAT outperforms other migration methods, achieving the highest diagnostic accuracy in case studies on three benchmark datasets: CWRU bearing dataset, SQ bearing dataset, Jiangnan University bearing dataset, and a real industrial dataset: Axial Fans fault dataset. The visualization results show that the model effectively extracts domain-divisible and domain-invariant features, exhibiting research prospects and application potential. The code library is available at https://github.com/MrTree777/GRF-GAT. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Intelligent Manufacturing is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)