Faults diagnosis of induction machine by using feed-forward neural networks and genetic algorithms

Abstract

We present the results of our investigation in the use of the multilayer feed-forward artificial neural networks (ANNs) and genetic algorithms (GAs) for fault diagnosis of induction machine. ANNs are used effectively to determine the classification of the faults of induction machine tested at different loads and at different frequencies. The novelty in this work is that proposed methodology is tested experimentally on four 4kW/1500rpm induction machines, with three current source frequencies (25,40,50)Hz on six different loads. The obtained results provide a satisfactory level of accuracy.