Neural approach for bearing fault classification in induction motors by using motor current and voltage

The induction motor is considered one of the most important elements in manufacturing processes. The use of strategies based on intelligent systems capable to classify the presence or absence of failures and also to determine its origin for the diagnosis and faults prediction is widely investigated in three phase induction motors. Thus, the aim of this paper is to present a methodology of bearing failures classification based on artificial neural networks, by using voltage and electric currents values in the time domain. Experimental results collected at real industrial process are presented to validate this proposal.