Application of PCA and SVM in Fault Detection and Diagnosis of Bearings with Varying Speed

Vibration analysis is widely used as an efficient condition monitoring (CM) tool for rotating machines in various industries. Fault detection and diagnosis (FDD) models play important role the development of any CM system. The use artificial intelligence (AI) has since gained recognition fault systems. In this paper, a combination principal component (PCA) which reducing data dimensionality, support vector machine (SVM) adopted classification to carry out faults bearings using vibrations. diagnostic feature design learning toolbox MATLAB are develop features train models, respectively. Real from Mendeley depository test evaluate models. Model training carried with varying speeds representing different conditions bearing making it similar approaches involving SVM. choice proves that SVM can be able classify consideration operating speeds. Results have shown PCA effective under such 97.4% accuracy was achieved. result implies implemented industrial setups where variable occur both intentionally or nonintentionally. Furthermore, method differentiate between compounding at times. confusion matrix further quality trained model. Future work will focus on prognosis well model other than faults.