A Current Noise Cancellation Method Based on Integration of Stator Synchronized Currents for Bearing Fault Diagnosis

Abstract

In bearing fault diagnosis using motor stator currents, anything unrelated to the fault characteristic frequencies is considered "noise". Noise cancellation provides a better estimation of the bearing fault signal. This paper aims to develop a noise cancellation method for a better estimation of the bearing fault signal in induction motors using three current signals, rather than just one. The paper proposes a method of time shifting in two phases of the three-phase current and integrating them with the third phase current as a noise cancellation method. The proposed method efficiently diagnoses bearing fault characteristic frequencies in spectral analysis. Simulation and experimental results confirm the effectiveness of the proposed method in detecting outer/inner raceway and ball bearing faults.