Data-driven Diagnostics for Electric Traction Systems: A Study of Induction Motor

Abstract

The topic of electric traction has garnered substantial attention in recent decades and continues to be the subject of intensive research efforts aimed at gaining market dominance, particularly with regards to traction motors. To ensure the reliable operation of electric motors in electric vehicles, it is crucial to incorporate on-board diagnostic and prognostic apparatus. Induction machines, which are widely utilized in electric traction systems, are of particular interest in this regard. This study introduces a novel diagnostic and detection technique for electric motor faults under random load conditions, utilizing an artificial intelligence approach. The technique was tested using a dataset comprised of both healthy state and malfunctioning state of induction motor, for which transient current, voltage, and vibration signals were recorded during motor launching and steady-state operation. The time-domain current data was processed using Fast Fourier Transform, resulting in frequency-domain data, which was then preprocessed prior to employing a supervised machine learning methodology to establish a diagnostic model that can determine the normal or abnormal functioning of the motor.