Knowledge Distillation and Enhanced Subdomain Adaptation Using Graph Convolutional Network for Resource-Constrained Bearing Fault Diagnosis

2025-01-13

AI Summary

Addressing the challenges of label scarcity, domain distribution shift, and resource constraints for edge deployment in bearing fault diagnosis under varying operating conditions, this paper proposes a progressive knowledge distillation framework. A GCN-ARMA model serves as the teacher, transferring knowledge to a lightweight student model. We introduce an Enhanced Local Maximum Mean Squared Difference (ELMMSD) metric—integrating mean/variance statistics in RKHS space with prior label probabilities—to enhance subdomain alignment robustness and clustering separability. Evaluated on CWRU and JNU datasets, our method achieves state-of-the-art diagnostic accuracy while reducing inference latency by over 60%. Ablation studies validate the effectiveness of each component, demonstrating significant improvements in cross-condition generalization and edge deployment adaptability.

Technology Category

• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Bearing fault diagnosis under varying working conditions faces challenges, including a lack of labeled data, distribution discrepancies, and resource constraints. To address these issues, we propose a progressive knowledge distillation framework that transfers knowledge from a complex teacher model, utilizing a Graph Convolutional Network (GCN) with Autoregressive moving average (ARMA) filters, to a compact and efficient student model. To mitigate distribution discrepancies and labeling uncertainty, we introduce Enhanced Local Maximum Mean Squared Discrepancy (ELMMSD), which leverages mean and variance statistics in the Reproducing Kernel Hilbert Space (RKHS) and incorporates a priori probability distributions between labels. This approach increases the distance between clustering centers, bridges subdomain gaps, and enhances subdomain alignment reliability. Experimental results on benchmark datasets (CWRU and JNU) demonstrate that the proposed method achieves superior diagnostic accuracy while significantly reducing computational costs. Comprehensive ablation studies validate the effectiveness of each component, highlighting the robustness and adaptability of the approach across diverse working conditions.

Problem

• Bearing Fault Diagnosis
• Resource Constraints
• Data Imbalance

Innovation

• Graph Convolutional Networks
• ELMMSD Method
• Bearing Fault Diagnosis