Early Fault Detection in Paper Machine Motors Using Machine Learning

Abstract

This research addresses the application of a neural network as a tool for early fault detection in the motors of a paper machine under a simulated environment. It proposes the analysis of variables from a torque control loop. The data for training and validating the model is obtained through the simulation of Direct Torque Control (DTC) of an AC motor in Simscape within Simulink. Both normal and faulty operating modes are considered. Under these two scenarios, various speed setpoints are configured, and the necessary data for training the developed model is collected