基于脉冲模式优化辅助的模型预测控制用于碳化硅永磁同步电机驱动以同时降低共模和差模电压

With the application of silicon carbide (SiC) power devices in permanent magnet synchronous motor (PMSM) drive systems, the voltage overshoot issue on the motor side becomes more prominent. Within the model predictive control (MPC) framework, the common-model voltage (CMV) reduction has been considered in some strategies. However, the reduction of differential-model voltage (DMV) for SiC-based PMSM drives under MPC has not been adequately investigated. Thus, this article proposes an improved MPC to simultaneously reduce CMV and DMV for SiC-based PMSM drives. In order to avoid additional hardware configurations, this work starts from the perspective of voltage vector pulse pattern optimization. In the proposed method, two groups of pulse patterns are designed according to the modulation index in which the modulation region is fully taken into account. Here, the reduction of both CMV and DMV spikes is looked after considering the dead time effect. In addition, with a well-tailored set of principles for the switching within and between the pulse patterns, the torque ripple can be further reduced without increase in switching frequency or computational burden. The proposed method is evaluated on a SiC-based PMSM drive setup and compared with the existing CMV-reduction MPC to demonstrate the effectiveness.