Model Predictive Torque Control of NPC Three-Level Inverter for Induction Motor Based on Generalized Two-Vectors Considering Neutral-Point Potential Balance

In order to maintain the neutral point potential balance and enhance the output current quality of the Neutral Point Clamped (NPC) three-level inverter, a novel approach called Neutral Point Balance based on Generalized Two-Vector Model Predictive Torque Control (NPB-GTV-MPTC) is proposed in this paper. The NPB-GTV-MPTC method evaluates vector combinations and their durations using a predefined cost function, aiming to achieve global minimization of torque error and optimal current quality. It effectively addresses the issue of neutral point potential imbalance in the NPC three-level topology. The effectiveness of the proposed NPB-GTV-MPTC is verified through a comparison of the inverter's output current quality, the motor's electromagnetic torque ripple, and motor speed response speed. The results demonstrate that NPB-GTV-MPTC outperforms the conventional GTV-MPTC approach, achieving superior performance in the considered aspects.