Diagnosis of Stator Asymmetry in Permanent Magnet Synchronous Motors Using Negative Sequence Current

In this paper, a new fault diagnosis approach is developed for stator asymmetry detection and localization in three-phase permanent magnet synchronous motors. Based on spectral analysis, Motor current signature analysis was extensively adopted for stator fault detection, but the same approach fails in identifying the faulty phase which causes the stator asymmetry of the motor. In the present contribution, a new approach is presented for stator fault detection and localization, which is derived from the polar coordinate of the negative-sequence current. A Neuro-fuzzy inference system is adopted for optimizing the fault localization process. The performances of the proposed approach are evaluated by means of numerical simulations, corresponding to a 7.5kW three-phase permanent magnet synchronous motor controlled by a conventional field-oriented control scheme.