Explainable fault and severity classification for rolling element bearings using Kolmogorov-Arnold networks

2024-12-02

arXiv.org

AI Summary

Rolling bearing faults are a primary cause of industrial system downtime and safety incidents, necessitating highly accurate and interpretable intelligent diagnostic methods. This paper introduces the Kolmogorov–Arnold Network (KAN) to bearing fault diagnosis for the first time, proposing a lightweight end-to-end framework that jointly classifies fault types and severity levels. The method integrates automatic feature selection, symbolic activation function modeling, and feature attribution analysis—eliminating black-box architectures in favor of explicit, differentiable functional representations to ensure intrinsic interpretability. It supports unified modeling of diverse mechanical faults, including imbalanced and misalignment conditions. Evaluated on two benchmark bearing datasets, the approach achieves 100% F1-score for fault detection and attains 100% accuracy across most fault-type and severity-level classification tasks. With significantly reduced parameter count, the model enables real-time deployment while preserving diagnostic transparency and generalizability.

Technology Category

• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Rolling element bearings are critical components of rotating machinery, with their performance directly influencing the efficiency and reliability of industrial systems. At the same time, bearing faults are a leading cause of machinery failures, often resulting in costly downtime, reduced productivity, and, in extreme cases, catastrophic damage. This study presents a methodology that utilizes Kolmogorov-Arnold Networks to address these challenges through automatic feature selection, hyperparameter tuning and interpretable fault analysis within a unified framework. By training shallow network architectures and minimizing the number of selected features, the framework produces lightweight models that deliver explainable results through feature attribution and symbolic representations of their activation functions. Validated on two widely recognized datasets for bearing fault diagnosis, the framework achieved perfect F1-Scores for fault detection and high performance in fault and severity classification tasks, including 100% F1-Scores in most cases. Notably, it demonstrated adaptability by handling diverse fault types, such as imbalance and misalignment, within the same dataset. The symbolic representations enhanced model interpretability, while feature attribution offered insights into the optimal feature types or signals for each studied task. These results highlight the framework's potential for practical applications, such as real-time machinery monitoring, and for scientific research requiring efficient and explainable models.

Problem

• Classify bearing faults and severity using interpretable neural networks
• Automate feature selection and hyperparameter tuning for fault diagnosis
• Provide explainable results through feature attribution and symbolic representations

Innovation

• Uses Kolmogorov-Arnold Networks for bearing fault classification
• Implements automatic feature selection and hyperparameter tuning
• Produces explainable models through feature attribution and symbolic representations