Physics-Informed Multimodal Bearing Fault Classification under Variable Operating Conditions using Transfer Learning

2025-08-10

AI Summary

To address the insufficient robustness and interpretability of bearing fault classification under varying operating conditions, this paper proposes a physics-informed multimodal CNN framework. Methodologically, it fuses vibration and current signals and introduces a physics-based loss function incorporating prior knowledge of bearing fault characteristic frequencies (BPFO/BPFI) to enforce mechanistic consistency in model outputs. A late-fusion architecture combined with three transfer learning strategies—Target-Specific Fine-Tuning (TSFT), Label-Adaptive Selection (LAS), and Hybrid Feature Regularization (HFR)—enhances cross-condition generalization. The key contribution lies in embedding fault characteristic frequencies directly into the loss function, thereby unifying interpretable modeling with physical consistency constraints. Evaluated on the Paderborn and KAIST datasets, the framework achieves 98% classification accuracy, with statistically significant reductions in misclassification rates (p < 0.01); among the transfer strategies, LAS jointly optimized with the physics-based loss yields the best performance.

Technology Category

• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

Accurate and interpretable bearing fault classification is critical for ensuring the reliability of rotating machinery, particularly under variable operating conditions where domain shifts can significantly degrade model performance. This study proposes a physics-informed multimodal convolutional neural network (CNN) with a late fusion architecture, integrating vibration and motor current signals alongside a dedicated physics-based feature extraction branch. The model incorporates a novel physics-informed loss function that penalizes physically implausible predictions based on characteristic bearing fault frequencies - Ball Pass Frequency Outer (BPFO) and Ball Pass Frequency Inner (BPFI) - derived from bearing geometry and shaft speed. Comprehensive experiments on the Paderborn University dataset demonstrate that the proposed physics-informed approach consistently outperforms a non-physics-informed baseline, achieving higher accuracy, reduced false classifications, and improved robustness across multiple data splits. To address performance degradation under unseen operating conditions, three transfer learning (TL) strategies - Target-Specific Fine-Tuning (TSFT), Layer-Wise Adaptation Strategy (LAS), and Hybrid Feature Reuse (HFR) - are evaluated. Results show that LAS yields the best generalization, with additional performance gains when combined with physics-informed modeling. Validation on the KAIST bearing dataset confirms the framework's cross-dataset applicability, achieving up to 98 percent accuracy. Statistical hypothesis testing further verifies significant improvements (p < 0.01) in classification performance. The proposed framework demonstrates the potential of integrating domain knowledge with data-driven learning to achieve robust, interpretable, and generalizable fault diagnosis for real-world industrial applications.

Problem

• Classify bearing faults accurately under variable conditions
• Integrate physics knowledge to improve model robustness
• Apply transfer learning for unseen operating conditions

Innovation

• Physics-informed multimodal CNN with late fusion
• Novel physics-informed loss function for fault frequencies
• Transfer learning strategies for unseen conditions