Enhancing three-phase induction motor reliability with health index and artificial intelligence-driven predictive maintenance

Abstract

The aim of this work is to assist in the maintenance of three-phase induction motors by creating a health index for this equipment. The proposed approach is based on power quality concepts, the creation of an algebraic algorithm to determine the health index and the use of artificial intelligence algorithms for modelling time series, such as Autoregressive Integrated Moving Average and Facebook Prophet, to predict the future health of the motor based on its historical data. The use of historical data makes it possible to anticipate potential failures and guide predictive maintenance strategies, helping to reduce costs and minimize unplanned downtime. The study examines various causes of failure in three-phase induction motors, analysing some of the most recurrent failures, their implications and the resulting impacts on the performance of the three-phase induction motor.