Position Fault Detection and Failover Method for UAM PMSM Control

A strict fault tolerance is required in urban air mobility (UAM). Position sensor is one of major maintenance target in permanent magnet synchronous motor (PMSM), since it is under hostile environments. As the sensor backup, sensorless algorithms are often used. In this work, sensorless algorithm and resolver work in cooperative and surveilling mode to each other. Based on the PMSM model, Runge-Kutta method is utilized to predict currents under two different position. These currents are compared with the real current measurements, and the result points to the angle suspected of being defective. Simultaneously, the deviation between the resolver sensor and the sensorless angle is always monitored. As this deviation, i.e. the angle error, increases, the weight of the suspected angle continuously decreases. Also, the fault is flagged when corresponding error exceeds a threshold value. In this work, a smooth transition is achieved thanks to logistic function-based weighting and sensor fusion The proposed method is verified through simulations and experiments.