FAULT DIAGNOSIS OF INDUCTION MOTOR USING ENVELOPE SIGNALS & POWER SPECTRAL DENSITY

Abstract

This paper proposes a novel and easy method for diagnosing the broken bar fault in induction motor. Discrete wavelet transform (DWT) along with Power Spectral Density (PSD) is used on the envelope of the stator current signal, a new dimension on fault diagnosis. The severity of the faults is also investigated. DWT is used for decomposing the signals into wavelet coefficients and PSD is used for calculating the energy of the wavelet coefficients. The energy of the higher level detail coefficients provides better fault diagnosis. The proposed method shows reliability in fault diagnosis and effectiveness in fault severity identification.