Machine Learning in the Prediction of the Insulation Temperature of Winding of Inductions Motors

Abstract

The operation of a motor with a temperature above the limit of its thermal class can significantly reduce the useful life of the winding insulation and consequently the useful life of the motor itself. With the industrial communication networks allied to the digital transformation process, it was possible to collect a series of data on the operation and operation of the induction motors, directing them to Machine Learning algorithms to produce results such as the prediction of the insulation temperature of the winding of a three-phase induction motor. This prediction process prompts maintenance departments to act before the temperature value exceeds a certain threshold.