Aug 2026· ISA transactions· 0 citations· 41 references
Medicine
TL;DR
A Structure-Preserving Constrained Smoothing Model (SPCSM) is proposed for partial cross-domain fault diagnosis, which is validated under time-varying speed conditions and outperforming several domain adaptation models.
Abstract
Fault diagnosis based on partial domain adaptation (PDA) has gained extensive attention, since it allows mismatched label sets between source and target domains and is thus more consistent with practical engineering applications. However, existing methods exhibit limited diagnostic performance under time-varying speed conditions. In this study, we propose a Structure-Preserving Constrained Smoothing Model (SPCSM) for partial cross-domain fault diagnosis, which is validated under time-varying speed conditions. By integrating the intrinsic geometric structure of the target domain with the category structure derived from the source domain, the model effectively enhances its smoothness, thereby mitigating the adverse effects of speed fluctuations on diagnostic performance. Experimental results demonstrate that the proposed SPCSM achieves an average accuracy of 98.06% on single-device tasks and 89.50% on cross-device tasks, outperforming several domain adaptation models.
Validation on multiple public bearing datasets demonstrates that IGDATL achieves higher accuracy and superior domain generalization performance, providing an innovative and practical solution for cross-domain intelligent diagnosis of complex industrial systems.
Xi-Yu Yang, You-Chao Sun, Jun-Fa Li· Proceedings of the Instituti...· 0 citations
A novel and robust Source-Free Domain Adaptation (SFDA) framework equipped with post-hoc explainability for bearing fault diagnosis is proposed, and a post-hoc visualization mechanism is incorporated to explain the model's decision-making process, enhancing the transparency and credibility of the diagnosis for practica...
Cheng-Hao Yan, Dong-Sheng Liu, Tong Wu et al.· Measurement science and tech...· 0 citations
Cross-domain fault diagnosis plays an important role in industrial production. Current methods are able to achieve promising performance when both source and target domains are available at training time. However, when oriented toward fault diagnosis where the target domain is unseen, they often fail to effectively ext...
Huikai Shao, Dong-Dong Jing, Ning Xia et al.· IEEE Transactions on Reliabi...· 0 citations
An explainable domain-adaptive CNN–Transformer framework for unsupervised bidirectional cross-domain bearing fault diagnosis using the Case Western Reserve University (CWRU) and Paderborn University (PU) datasets, which confirms the complementary contributions of convolutional feature extraction, Transformer-based temp...
Muhammad Javed, Su-Hang Ding, Hong-Xia Yan et al.· Electronics· 0 citations
PDAPN first employs episodic prototypical learning to construct stable and discriminative class representations from scarce labeled source-domain samples, and a maximum of cosine similarity-based local partial alignment mechanism is introduced to identify target-compatible source-domain anchors and reduce negative tran...
Zhi-Tong Liu, Rengen Wang, Liu Cheng et al.· Structural Health Monitoring· 0 citations
Intelligent fault diagnosis (IFD) models that rely on one-shot learning often struggle in dynamic mechanical systems because they exhibit limited knowledge stability and high computational cost under incremental learning (IL). To address the trade-off between robustness and efficiency, this paper proposes a Multi-Scale...
Sheng-Jie Zhang, Bao-Yi Xu, Ze-Yun Yang et al.· Neural Networks· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
