Symmetrical components features analysis for inter-turn fault detection in permanent magnet synchronous machines

Abstract The implementation of the Permanent Magnet Synchronous Machines (PMSMs) has been increasing in recent years. They play a central role in the advancement of electric vehicles, the expansion of wind energy generation, and the improvement of industrial efficiency. However, detecting inter-turn short-circuit (ITSC) faults remains a challenge, especially in their incipient stage, as these faults often begin with the loss of only a few coil turns. If undiagnosed, such faults can spread quickly, drastically shortening the lifespan of the machine. This paper presents a detailed analysis of ITSC fault detection in PMSMs based on the inspection of symmetrical components features of the stator currents and/or voltagesh. Features such as positive, negative, and zero-sequence phasor modules, angles, and their relative variations are explored. The study is conducted by analyzing the data obtained from a digital model and by validating them through experimental measurements in a laboratory using a 3.5 kW surface-mounted PMSM. Particular attention is given to identifying the most relevant and accurate diagnostic signature measurements. Results for steady-state diagnostics show that the negative angle variation relative to the positive sequence allows a fast ITSC fault detection regardless of the operating point of the machine, even for high fault resistances.