A modulated triple vectors model predictive controllers for PMSM drives based on voltage reference position selection

• Triple voltage vector approach with an optimized duty cycle modulation scheme. • Cost function minimization for optimal current prediction. • The duty ratio for each voltage vector is calculated online for smooth modulation. • Model predictive speed controller with current constraints for dynamic speed improvement. • Reduce the computational complexity while minimizing current and torque ripples. This study introduces a modulated triple vectors model predictive current controller. The proposed method employs a triple voltage vector approach with an optimized duty cycle modulation scheme to improve current prediction accuracy while minimizing torque and current ripples. The control algorithm utilizes cost function minimization for eight possible voltage vectors to determine the optimal current prediction and properly selects voltage vector combination, thereby enhancing dynamic response and steady-state performance. By properly selecting voltage vectors based on the reference position in a stationary reference frame, the proposed method reduces computational complexity compared to conventional single vector and dual vector approaches. In addition, a model predictive speed controller with constraints is implemented to improve the dynamic speed controller. Experimental results confirm the advantages of the proposed method by significantly reducing total harmonic distortion and torque ripple, which are 5,26% and 0,105 N.m, respectively. Additionally, the proposed method exhibits improved robustness under different speed and load disturbance conditions, making this proposed method become a possible solution for high-performance permanent magnet synchronous motor drive applications.