Online Multiparameter Identification of Surface-Mounted PMSM Considering Inverter Disturbance Voltage

The disturbance voltage of voltage-source inverter operation exerts significant influences on the accuracy of motor parameter identification. In this paper, an accurate model is established for estimating the disturbance voltage that includes current harmonics and their derivative. The conclusion is drawn that the derivative of the d-axis current harmonics plays a major role in causing estimation error of the disturbance voltage. Based on this, two methods are proposed to reduce the estimation error of the disturbance voltage. One employs an iterative learning controller in the d-axis current control loop to suppress the harmonics of d-axis current, thus reducing their influence on the estimation of the disturbance voltage. The second estimates the disturbance voltage on the basis of the average value of the variables in q-axis equation of the motor, so that the estimation result is immune to current harmonics. Motor parameter identification accuracy is consequently improved by both.