A Double Vector Model Predictive Torque Control Method Based on Geometrical Solution for SPMSM Drive in Full Modulation Range

Abstract

To avoid the complex process of the weighting factor designing in model predictive torque control (MPTC) of surface permanent-magnet synchronous motor (SPMSM), a novel double vector MPTC without weighting factor is proposed in this article. First, the cost function of the torque and stator flux is converted into the voltage function in the two-phase synchronous stator frame. Meanwhile, the coordinate expression of an optimal reference vector, which can simultaneously satisfy the deadbeat condition of torque and flux, is derived by the proposed geometrical method. Then, to minimize the distance from the synthesizing vector to the reference vector, the zones of the linear modulation and overmodulation are divided into several parts. Accordingly, the principle of the double vector selection and the duty cycle calculation are presented in detail. Meanwhile, to improve the robustness, a detailed robustness analysis is conducted and an effective parameters identification method is proposed. Finally, the experimental verification is carried out on a 1 kW SPMSM drive system. Compared with two existing double vector MPTC methods, the proposed method can reduce the THD of the stator current, torque ripple, and stator flux ripple. Meanwhile, a faster dynamic response can be obtained by the proposed MPTC method.