Predictive controller design for a sensorless IPMSM speed control system

This paper proposes a predictive controller design for a sensorless interior permanent magnet synchronous motor drive system. The controllers include a current-loop predictive controller and a speed-loop predictive controller with an external load estimator. In addition, the estimated error of the rotor position estimator is reduced due to the using of the on-line tuning stator resistance and q-axis inductance. By using the proposed control algorithms, the closed-loop sensorless interior permanent magnet synchronous motor drive system can achieve fast transient responses, satisfactory load disturbance responses, and good tracking responses. A digital signal processor, TMS 320F2812, is used to perform the rotor position estimating algorithm and the predictive control algorithm. As a result, the hardware is very simple. Experimental results can validate the theoretical analysis and show that the predictive controllers have better performance than the PI controllers.