Presentation Details
  1. Home
  2. Presentation List
  3. Presentation Details
A Generalize Mechanical Fault Diagnosis Method Based on Enhanced Meta Learning
ID：131 View Protection：ATTENDEE Updated Time：2025-11-10 15:52:08 Hits：118 Poster Presentation

Start Time：Pending（Asia/Shanghai）

Duration：Pending

Session：No Session »

No files

Abstract
Mechanical fault diagnosis faces dynamic and complex fault patterns and environmental variations, requiring diagnostic models with high generalization capability. To mitigate the data distribution shift under different operating conditions and address the class imbalance caused by the high cost of data collection, this paper proposes an enhanced meta learning framework. From a gradient representation perspective, this framework learns generalized boundaries across all diagnostic tasks by coordinating inter-domain and inter-class gradients, thereby improving the recognition of unknown classes under imbalanced conditions. Furthermore, based on this enhanced meta-learning framework, a joint learning paradigm involving open- and closed-set classifiers is developed to balance the decision boundaries between known and unknown classes, enabling the model to quickly adapt to unfamiliar domains. The superior performance of the proposed framework is validated on the publicly recognized HIT bearing dataset.
Keywords
class imbalance, fault diagnosis, meta learning, rotating machinery
Speaker
Yan Ge
senior engineer China Institute of Marine Technology & Economy

Submission Author
Yang Song China State Shipbuilding Corporation Limited
Qingyuan Cao Beijing Institute of Structure and Environment Engineering
Yan Ge China Institute of Marine Technology & Economy
Submit Comment
Verify Code Change Another
All Comments
Important Date

  • Conference Date

    Nov 21

    2025

    to

    Nov 23

    2025

  • Oct 20 2025

    Draft paper submission deadline

  • Dec 08 2025

    Registration deadline

Sponsored By
IEEE Instrumentation and Measurement Society
South China University of Technology
Organized By
South China University of Technology
Contact Information
Login to view full contact details
Previous Conferences