Journal
RELIABILITY ENGINEERING & SYSTEM SAFETY
Volume 230, Issue -, Pages -Publisher
ELSEVIER SCI LTD
DOI: 10.1016/j.ress.2022.108890
Keywords
Fault diagnosis; Bearings; Imbalanced domain adaptation; Label shift; Domain shift
Ask authors/readers for more resources
The success of deep learning and transfer learning has expanded the scope of fault diagnosis, especially in improving diagnosis accuracy under multiple working conditions. However, most existing approaches do not account for the diversity of fault mode distributions and weaken the generalization to imbalanced domain adaptation scenarios. This work proposes a novel deep imbalanced domain adaptation framework for fault diagnosis, which overcomes class-imbalanced label shift and improves cross-domain generalization for IDA tasks.
The tremendous success of deep learning and transfer learning broadened the scope of fault diagnosis, especially the latter further improved the diagnosis accuracy under multiple working conditions. However, most existing attempts assume that label distribution is domain-invariant despite taking into account the different feature distributions. This does not accommodate the diversity of fault mode distributions under different operating conditions and weakens the generalization to imbalanced domain adaptation (IDA) scenar-ios. Therefore, this work proposed a novel deep imbalanced domain adaptation (DIDA) framework for fault diagnosis of bearings, aiming at the challenging scenario where feature shift and label shift exist simultaneously under different working conditions. Specifically, DIDA overcomes the class-imbalanced label shift and achieves a fine-grained latent space matching by cost-sensitive learning and categorical alignment. Besides, margin loss regularization is introduced to further optimize classification boundaries and improve cross-domain generalization for IDA fault diagnosis tasks. Finally, we simulated the IDA protocols on experimental datasets and conducted case studies under multiple working conditions, thus validating the effectiveness and superiority of the proposed framework.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available