An Effective Multivector Model Predictive Current Control for PMSM Drive Based on Low-Complexity Voltage Vector Preselection

Model predictive current control (MPCC) is an advanced control strategy for nonlinear constrained systems that demand quick dynamic response. An effective multivector MPCC technique based on a low-complexity voltage vector (VV) preselection is proposed in this article for the permanent magnet synchronous motor (PMSM) drive. The proposed scheme employs three vectors in a sample time as the single VV application leads to large fluctuations in torque. As the application of multiple VVs augments the complexity of the algorithm, the proposed scheme opts for the principle of current error minimization as a criterion for VV preselection. Hence, the proposed MPCC technique employs the previous sample VV and the motor rotation direction as parameters for preselection, thereby limiting the computations using a set of four VVs. Moreover, the application of multiple VVs is achieved using a gradient-based approach, selecting the first optimum VV from the preselected set and determining the suitable adjacent VV as the second optimum VV through voltage error gradients. Application times are then precisely determined using the deadbeat concept, effectively reducing the steady-state torque fluctuations without compromising dynamic response. Furthermore, the proposed control algorithm is comprehensively compared with basic MPCC (BMPCC) and recent literature to ensure its effectiveness.