Fault Prognosis of Induction Motor Using Multi Resolution Current Signature Analysis

There are various methods for the condition monitoring and this paper focuses on the multi resolution current signature analysis for fault prediction of induction motors. Variable frequency drives-based induction motors are used widely in industries. Monitoring the health of the motors is of great importance to reduce downtime and increase productivity. The multi resolution coefficients features from current signal are extracted using empirical wavelet transform. The extracted features are fed as input to artificial neural network to do prognosis on the data obtained for finding the condition of the motor. Hall Effect based system is used to measure the current signal and the features are extracted and trained to predict the condition of system using MATLAB in real time. The experimental findings reveal that the suggested technique achieves better accuracy in induction motor fault prognosis.