Application of statistical neuronal networks for diagnostics of induction machine rotor faults

Induction machines are extensively used in industries and are subject to unexpected breakdowns. It is necessary, therefore, to prevent them from such breakdown through the maintenance that works according to a well-trained planning. A considerable number of diagnosis techniques have been used such as Motor Current Signature Analysis (MCSA), Axial Flux Monitoring and Vibration Monitoring. This paper shows the effectiveness of the artificial neuronal network (radial basis function neuronal network and the probabilistic neuronal network) basis on MCSA for rotor faults diagnosis.