Torque ripple RMS minimization in model predictive torque control of PMSM drives

Abstract

This paper proposes an improved model predictive torque control (MPTC) with duty ratio optimization for three-phase AC permanent magnet synchronous motor (PMSM) drives. By predicting the evolution of variables concerned using system model, the best voltage vector satisfying the torque and flux command can be selected. As a result, MPC has a fast dynamic response and good steady-state response. Despite the merits above, the traditional MPC still presents some torque and flux ripples due to the use of single voltage vector during the whole control period. To further improve the steady state performance of conventional MPTC, this paper introduces the concept of duty cycle optimization in MPTC by dividing the control period into two intervals. The non-zero vector obtained from conventional MPTC is applied for a fraction of control period based on the principle of torque ripple minimization, while the rest of time is allocated for a null vector. Simulation results are presented to validate the effectiveness of this method.