Bearing Fault Detection in Induction Motors Using Line Currents

This paper focuses on the development of a bearing fault detection model for induction motors using line currents. The graphical and numerical analysis of the model has been developed using the park's vector approach and envelope based on the Hilbert transform. The proposed model has been evaluated on currents measured from eight different types of induction motors. The graphical results from the Concordia pattern between d- and q-components of stator currents show that healthy bearing behaviour is circular compared to the faulty bearing's elliptical. The numerical results show that the minimum and maximum envelope of d- and q-components of stator currents is more significant than one. The sum of Kurtosis for the envelope of d- and q-components of stator currents is less than 5.0.