Effective Electrical Fault Diagnosis Using Gabor Transform of Motor Input Current After Spectral Subtraction.

In the present work, it is proposed to use stator current spectrum analysis by Gabor transform (GT) to identify electrical defects in induction motors. Current-based fault detection plays an important role in the proposed work as it overcomes the major pitfalls associated with other conventional sensor-based monitoring. A Poly Phase Induction Motor (IM) of 2 HP, 415 ± 10 % V, 50 Hz is taken for online testing of proposed fault detection at variable speed regime. At varying load situations, the motor's input current is gathered for various fault conditions. After performing frequency spectral subtraction, advanced GT offers signal processing on the collected stator current. The GT coefficients are used for further analysis to estimate and localize the fault. Fault indexing parameter (FIP) is proposed for fault severity estimation. The data acquisition is carried out in LabVIEW environment and processed by MATLAB programming. [ABSTRACT FROM AUTHOR]