传统模型预测控制(MPC)采用电压源逆变器（VSI）生成的全部7个电压矢量作为备选电压矢量集合，造成较大的运算负担。不同条件下，模型预测转矩控制（MPTC）系统对7个电压矢量的实际利用率并不均衡。在考虑定子磁链扇区、磁链和转矩控制信号和转矩角条件下，分析了表贴式永磁同步电机(SPMSM)MPTC系统对7个备选电压矢量的实际利用情况，并基于电压矢量利用率规律，提出了可动态变化的备选电压矢量集合。仿真结果表明与传统备选电压矢量集合相比，所提备选电压矢量集合可在保持控制性能基本相当的同时，减小备选电压矢量个数，有效提高系统实时性能。

Traditional model predictive control (MPC) uses all 7 voltage vectors generated by voltage source inverter (VSI) as candidate voltage vectors, which results in heavy calculation burden. But traditional model predictive torque control (MPTC) does not utilize 7 voltage vectors equally under some conditions. The MPTC system for surface permanent magnet synchronous motor (SPMSM) is built. Considering different stator flux sectors, flux and torque control commands and torque angles, the utilization rate of 7 voltage vectors is studied. Based on the principle of utilization rate of voltage vectors, simplified candidate voltage vector sets are proposed which can dynamically adjust candidate voltage vectors. Simulation results show that compared with traditional candidate voltage vectors sets, the proposed sets maintain the performance equivalently and decrease the number of candidate voltage vectors, and therefore can improve the realtime performance of the system.

国家自然科学基金项目(51207012)；陕西省工业科技攻关项目(2016GY-069)；陕西省微特电机及驱动技术重点实验室开放基金项目(2013SSJ2002)；中央高校基本科研业务费专项资金项目（300102228201）