Sensorless Based Finite Control Set Model Predictive Current Control of PMSMs with PM Flux-linkage Immunity

Finite control set model predictive current control has fast dynamics and is simple to implement. However, it requires accurate machine parameters for the current prediction, any parameter mismatch will cause a current prediction error resulting in deteriorated control performance. Particularly, PM flux-linkage is commonly required by the prediction model and mismatch in PM flux-linkage will cause current prediction error. This paper proposes a finite control set model predictive current control with PM flux-linkage immunity based on back-EMF sensorless control. For the sensorless control based system, the back-EMF is estimated and regulated by the sensorless algorithm, which can be directly used in the prediction model. Thus, the knowledge of PM flux-linkage is not required in the prediction model. Moreover, by utilizing the estimated back-EMF, the proposed method has an enhanced robustness against resistance and inductance mismatches. The proposed method is verified by experiments and shows better control performances compared to the conventional method in terms of steady- and dynamic-states.