PMSM identification for automotive applications: Cancellation of position sensor errors

The rotor position plays a significant role in identification of motor parameters in Permanent Magnet Synchronous Machines (PMSMs). This paper presents a new model, taking into account errors in measurement of rotor position, for PMSM. A Least Squares (LS) algorithm, based on this new model, is also presented. The proposed algorithm relies on currents, voltages and speed measurements, and identifies both motor parameters and the value of the position error. The method is experimentally implemented and the results demonstrate that the new proposed model and LS algorithm improve the identified resistance and dq-axis inductances, while the rotor flux is hardly affected by the position error. Moreover, a study of measurement uncertainties is conducted to establish confidence intervals on the identified parameters.