Abstract: Due to factors such as the structural design of the motor and control methods, brushless direct current motors (BLDCM) generates torque ripple during operation, causing the motor to vibrate during full-speed operation. In this paper, the causes of commutation torque ripple are analyzed, and an improved model predictive control method is proposed in combination with the duty cycle prediction method, which replaces the traditional current loop PI controller with duty cycle prediction, and uses the predicted duty cycle for the generation of PWM_ON_PWM modulation waves, which avoids the complex parameter settings of the traditional PI control and effectively reduces part of the torque fluctuation. In the process of current prediction, a dynamic feedback correction method is proposed to eliminate the prediction value offset problem caused by changes in motor operating parameters. Finally, simulation experiments verified the effectiveness of the proposed control strategy in reducing commutation torque ripple at different speeds, and significantly improved the control performance of the brushless DC motor control system.

    Key words: brushless direct current motor; torque ripple; model prediction; dynamic correction; duty cycle; cost function