A Light‐Computational Model Predictive Control With Current Enhancement for PMSMs Based on Discrete Space Vector Modulation

Although the model predictive control (MPC) with discrete space vector modulation (DSVM) can effectively mitigate current ripple, thus reducing torque ripple, the computational burden increases dramatically with candidate voltage vectors. In this paper, a light‐computational MPC with current enhancement for permanent magnet synchronous machines based on DSVM is proposed, aiming to decrease the computational complexity and further improve the performance of the MPC with DSVM. Different from the existing DSVM, the proposed DSVM synthesizes virtual voltage vectors using only four real voltage vectors. In addition, it is noteworthy that the duty cycles of three nonzero voltage vectors corresponding to the virtual voltage vector are also duty cycles of the inverter upper bridge arms, thus eliminating the inverter switching signal calculation. Based on the proposed DSVM strategy, a voltage vector preselection method is designed to reduce candidate voltage vectors by four. Thereafter, voltage vectors around the optimal voltage vector are extended to further enhance current performance. Experiments are conducted on a 2‐kW electric drive platform to verify the feasibility and effectiveness of the proposed method.