Detection and isolation of asymmetrical short-circuit faults in permanent-magnet synchronous machines

Phase-to-phase short-circuit faults in a permanent-magnet synchronous machines are quite dangerous because of the extremely high short-circuit transient current after fault. This paper presents a fault detection and identification diagnosis approach for asymmetrical faults in permanent-magnet synchronous machines, considering both inter-turn and phase-to-phase short-circuit faults, and employing both current- and voltage-based residual along with the cumulative-sum algorithm. The sum of positive- and negative-sequence current- and voltage-phase angles is used to identify the faulty phase(s); the two fault types are distinguished by defining different detection thresholds for the residuals. Simulations of phase-to-phase short-circuit with different short-circuit resistance have been carried to obtain the values of faulty phase indicators. Both the simulation results and the proposed diagnosis have been verified by practical experiments on a 2-kW test bench.