Diagnosis of stator short-circuit faults in an IPM synchronous machine using a space-vector pendulous oscillation method

Abstract

Experimental characterization of interior permanent magnet (IPM) synchronous machines experiencing winding short-circuit faults is necessary in developing diagnostics indices to be used in detecting short-circuit winding faults. Known fault indices include the current spectrum analysis, the negative-sequence components method, and the space-vector based pendulous oscillation method. This paper applies all three methods to experimental turn-to-turn short-circuit fault data acquired from a 3.5 hp IPM motor energized by a PWM voltage source inverter. The results are compared and analyzed, and since none of the three methods presented unambiguously demonstrate the existence of a winding short-circuit fault, the practical difficulties of short-circuit fault detection and mitigation are discussed in that light.