Electromechanical brake modelling and control: from PI to MPC

Abstract—The electromechanical brake (EMB) force control problem has been approached in prior work using cascaded proportional-integral (PI) control with embedded feedback loops to regulate clamp force, motor velocity, and motor current/torque. However, this is shown to provide limited performance for an EMB when faced with the challenges of actuator saturation, load-depen-dent friction, and nonlinear stiffness. There is a significant margin for improvement, and a modified control architecture is proposed using techniques of gain scheduling, friction compensation, and feedback linearization. A further improvement is then achieved by incorporating a model predictive control that better utilizes the available motor torque. Simulation and experimental results are presented to demonstrate the improvement in performance. Index Terms—Cascaded proportional-integral (PI) control, electromechanical brake (EMB) model, electromechanical brakes (EMBs), feedback linearization, friction compensation, gain scheduling, model predictive control (MPC). I