Abstract:
In industrial plants, engines are subject to failures that can cause production downtime causing
economic losses. Predictive monitoring identifies these failures and reduces maintenance and
operating costs. In this work, systems for monitoring and diagnosing faults in three-phase
induction motors were developed based on the IEC 60034 and ISO 20816 standards. Through a
pattern recognition neural network, the Operating Regions Diagnostic System classifies the region
of operation of the motor. The Rotor Broken Bar Diagnostic System monitors the machine’s state
through a recurring neural network of short-term long memory, identifying the number of broken
bars on the rotor and the operating zone according to ISO 20816. Finally, the Diagnosis Fuzzy
uses a neuro- fuzzy system to predict the motor reduction factor, predicting possible failures that
occur due to voltage imbalance.