An Autoencoder-Based Algorithm for Fault Detection of Rotating Machines, Suitable for Online Learning and Standalone Applications

Development of an autoencoder-based algorithm for fault detection in rotary machines is presented in this paper. As the cornerstone of any machine learning model, feature engineering is thoroughly studied to select the best set of features for the application. Then, an autoencoder architecture is employed to detect the anomalous behavior of the machine. A laboratory setup is designed and built to provide the training and test data and validate the proposed algorithm. It is also shown that the proposed method is well suited for online learning and real-time applications.