Improved Three Vector Model Predictive Torque Control of PMSM

To reduce the computational complexity of traditional model predictive torque control (MPTC) and improve the sensitivity of predictive control to disturbances, an improved three vector model predictive control strategy applied in permanent magnet synchronous motor (PMSM) is proposed. First, the principle of deadbeat synchronization between torque and flux linkage is adopted to reduce six candidate vectors in traditional torque prediction to two, and the cost function is designed to select the optimal voltage vector. In addition, disturbance observation compensation is introduced to compensate for the influence of load disturbance on the control performance of the predictive model. As experimental results show, the proposed three-vector model predictive torque control can obtain small torque ripple and current harmonics both in steady state and dynamic state.