Preliminary Detection of Bearing Faults using Shannon Entropy of Wavelet Coefficients

In this work bearing faults of a three phase induction motor are detected using the information contents of stator current. The proposed methodology is implemented by taking Wavelet transform of the three phase currents of the induction motor. Finally the diagnosis is performed by employing the Shannon Entropy of the detailed wavelet coefficients at the decided level of decomposition. The work discusses some of the commonly occurring faults in the motor and a brief overview of Wavelet Transform. For different parameters the systems information and experimental results are demonstrated applying MATLAB toolbox. As a result the proposed methodology could improve the accuracy and reliability in the detection of induction motor faults.