Multi-stage Series Model Predictive Torque Control for SPMSM Drives

Abstract

A multi-stage series model predictive torque control suitable for permanent magnet synchronous motors is proposed in this paper. Compared with the conventional MPTC, the weighting factor is eliminated and the switching frequency is reduced while the steady-state control performance is guaranteed. Firstly, the instantaneous power theory is introduced. The control of torque and flux linkage is transformed into double torque control of active torque and reactive torque by using the idea of instantaneous active power and reactive power. Then, the predicted value of double torque in the next $m$ moments is predicted by iterative operation ($m$-stage series method), and the cost function is formed by comparing with the extrapolated reference value stage by stage, the optimal voltage vector is selected gradually through the $m$ cost functions and applied to the converter. Finally, the effectiveness of the method is verified by experiments. The proposed method not only guarantees the steady-state control performance, but effectively also reduces the switching frequency.