A Three-Vector Fast Model Predictive Control Method for Steady State Performance Improvement

A fast three-vector model predictive control method based on an improved H8 inverter is proposed for the Permanent Magnet Synchronous Motor (PMSM) driven by a two-level three-phase voltage source inverter. The method aims to reduce torque ripple, lower total current harmonic distortion (THD), and common-mode voltage, thus improving motor steady-state performance. The proposed method first expands the candidate voltage vectors of the improved H8 inverter by introducing the zero vector into the candidate set, thereby reducing current harmonic content. Then, the method selects the optimal voltage vector online from six effective voltage vectors based on current prediction, computes the error vector angle between the predicted current and reference current, and chooses a second voltage vector to minimize the current error. A zero vector is used to adjust the output voltage amplitude, and a new method for calculating the duty cycle of the voltage vectors is designed, considering the impact of dead-time on common-mode voltage. The voltage vector switching sequence is centralized to avoid equivalent zero vectors caused by dead-time, improving system performance. Finally, simulations compare the proposed method with traditional two-vector model predictive control, three-vector model predictive control, and model predictive control with common-mode voltage suppression. Results show that the proposed method effectively prevents common-mode voltage spikes caused by dead-time, reduces torque and flux ripples, and minimizes current harmonic content while lowering computational complexity.