Machine learning algorithms for rotating machinery bearing fault diagnostics

Abstract

Abstract In present-day reducing the catastrophic failure and breakdowns of rotating machinery is quite important to achieve the desired quality of the product. Simultaneously to attend high production rate as required. Challenging tasks to achieve better machine performance, intelligent diagnosis of rotating machinery will play a vital role. Bearing faults are the main cause of the breakdown of any rotating machinery. In this paper comprehensive review of machine learning algorithms and techniques implemented for rotating machinery bearing faults are reviewed with respect to various parameters. Various methods of machine learning like Support vector machine (SVM), artificial neural network (ANN), decision tree (DT) and K-nearest neighbor (KNN), relevance vector machine (RVM), and support vector regression (SVR) are compared based on theoretical background and its industrial applications.