Research on model predictive control of magnetorheological impact buffer system considering driver modeling

Abstract The working environment of the magnetorheological (MR) damper under high- speed impacts is harsh, the uncertainties in the models of each system component, unmodeled parts, as well as internal and external disturbances, can significantly affect the buffering performance. This paper suggests a model predictive control (MPC) method for the MR impact buffer system that incorporates the driver model to enhance the control performance of the system. The Buck driver circuit was first accurately modeled mathematically, and then the driver’s hardware was fabricated. To confirm the effectiveness of the developed driver, experiments were conducted to test the driver’s response time, ripple, and linearity of output current. The state equations of the system are obtained by linking the Buck circuit mathematical model with the magnetic induction response characteristics equation, and the prediction model of the system is produced through discretization. The impacts of temperature, hysteresis disturbance, and driver model on the system’s control performance are examined using Simulink simulation. The results indicate that temperature, hysteresis disturbance, and driver model impacts increase the system’s peak damping force by 466.8 N and 724 N, respectively. The MR impact buffer system contains the driver model’s MPC simulation results, showing that the system’s peak damping force is 4534.9 N, which represents a reduction of 32.5% and 17.7% compared to the open-loop control and the proportional-integral-derivative (PID) control, respectively, and the control effect demonstrates a significant ‘platform effect,’ confirming the superiority and efficacy of the proposed control approach.