Online identification method of static & dynamic inductance of IPMSM for fine position sensorless control

Abstract

When position-speed sensorless control is used, the estimation error of the position often becomes large because of inductance value fluctuation due to the current. Therefore, this paper proposes a online parameter identification method to identify the static inductance and the dynamic inductance and magnet flux by using fixed trace algorithm. In the proposed method, the harmonic is superimposed into the motor voltage reference or motor current reference. As the result, the estimation error of the position is reduced by the proposed method because the identified inductance is reflected in the position-speed sensorless control. The validity of proposed method is confirmed by the numerical simulation results and the experiment results in this paper.