Robust model predictive control for PMSM drives based on current prediction error

Abstract

The model predictive current control (MPCC) uses the present sampling current to predict the next cycle current, and select the optimal voltage vector from seven voltage vectors, which will be acted on machine in next control cycle. However, the control performance of traditional MPCC is influenced by the accuracy of parameters. In order to solve the parameter sensitivity of the MPCC, a robust model predictive control method is proposed, which is based on current error that presents the difference between current predicted value and current actual value. Proposed method uses this current error to obtain the parameter information of inductance and flux linkage. The simulation and experiment indicate that proposed method can obtain accurate parameters under different operation conditions and control performance of the whole control system is excellent.