Rolling bearing fault diagnosis based on feature extraction and support vector machine

Abstract

Abstract Rolling bearing fault diagnosis is essential for maintaining the stable and safe operation of rotating machine systems. This paper addresses the critical challenge of rolling bearing fault diagnosis by proposing a “feature extraction-fault diagnosis” framework. First, feature extraction method is applied to extract features from both time- and frequency-domain. Then, these extracted features are fed into support vector machine (SVM) for fault classification. Finally, the proposed method is compared with several benchmark models on the Case Western Reserve University (CWRU) rolling bearing fault dataset to demonstrate its performance. The results of case study indicate that SVM model achieves accuracy of 99.17% in two cases, outperforming comparative models in feature extraction capability and classification precision. The results confirm the framework’s effectiveness in achieving high-precision fault diagnosis, offering a reliable solution for the maintenance and operation of rotating machine systems.