Rotor faults diagnosis using artificial neural networks and support vector machines

Abstract

Unbalance and misalignment are the commonly occurring faults in rotating mechanical systems. These faults are caused mainly due to improper installation or premature failure of the machine components. Detection and diagnosis of faults in rotating machinery is crucial for its optimal performance. In this study artificial neural networks (ANN) and support vector machine (SVM) techniques have been used to determine the effectiveness of statistical features for fault diagnosis in rotating mechanical system using healthy and faulty rotors. The vibration signature responses are obtained and analyzed for healthy shaft without disk (HSWD), healthy shaft with an unbalanced disk (HSWUD), centrally bent shaft without disk (CBSWD) and centrally bent shaft with an unbalanced disk (CBSWUD) with zero bow phase angle. Their predominant features were fed as input for training and testing ANN and SVM, whereas the relative efficiency of these techniques have been compared for classifying the faults in the test system. The study concludes that these machine learning algorithms can be used for fast and reliable diagnosis of rotor faults.