Induction motor rotor fault detection using Artificial Neural Network

Abstract

The present paper deals with the detection of broken rotor bar of an induction motor. The problem is approached through mathematical modeling of induction motor. Both the models, for healthy as well as faulty motor, are developed using MATLAB simulink. The model is used to simulate different conditions of fault with varying number of broken bars. Parameters like three-phase voltage, three-phase current and THD of all voltages and currents are acquired from the simulated model. The data thus generated is used to train Artificial Neural Network which diagnoses the condition of motor. The results obtained prove the effectiveness of proposed method.