Recent advances in fault diagnosis by Park's vector approach

This paper addresses recent advances in fault diagnosis by Park's Vector Approach, with special emphasis being given to the most recent developments concerning power converter fault diagnosis in variable speed ac drives. With the aim to demonstrate the suitability of such approach for power converters self-diagnostic, a detailed comparison of four algorithms intended for real-time diagnosis is also included. Their performance evaluation is accomplished by means of experimental results, analyzing some key features such as effectiveness, robustness against false alarms, detection speed, implementation effort, tuning effort and computational burden.