Fault detection in induction motors using Park's vector approach and wavelet analysis

In this paper, Park's vector analysis with wavelet packet transform has been proposed in order to detect bearing faults of induction motor. Park vector eliminates the main frequency of the original current measurement signal and increases the amplitude of the fault components. Employing wavelet packet transform, due to its ability in non-stationary signal analysis, improves the performance of the fault diagnosis approach. Signal energy is considered as a criterion to compare normal and faulty conditions. Experimental results show the superiority of the proposed method in fault diagnosis at early stages.