A Nonlinear State Observer for Sensorless Speed Control of IPMSM

Abstract

This paper presents a nonlinear state observer model for sensorless speed control of an interior permanent magnet synchronous motor (IPMSM) drive. First, an augmented state space model of the IPMSM and the system uncertainty dynamics is obtained in the d − q synchronously rotating reference frame. Then, the proposed nonlinear state estimator based on an extended Luenberger observer is constructed by using rotor position, rotor speed, q-axis stator current, and system uncertainty vector as the state variables. Hence, the estimated system parameter uncertainties can be used to dramatically improve the performance of the speed controller without any additional observer. The stability of the proposed observer at low and high speed is asymptotically guaranteed by Lyapunov stability analysis. In addition, the proposed model is simple to carry out, so it does not require a complex calculation. Simulation results obtained via a Matlab/Simulink software indicate that the proposed observer model can assure an excellent state estimation over a wide speed range even when the rotor speed or load torque changes abruptly as well as the difference in the initial values of real and estimated rotor position.