Model predictive regenerative braking control for lightweight electric vehicles with in-wheel motors

Abstract

This paper presents a nonlinear model predictive controller for regenerative braking control of lightweight electric vehicles equipped with in-wheel motors. In-wheel-motors-driven electric vehicles possess significant advantages such as actuation flexibilities, torque control precision, and energy recovery improvement by direct regenerative braking control. The proposed controller not only improves the regenerative braking energy recovery by determining the front and rear braking torques independently but also prevents wheel locks during deceleration when the tire–road friction coefficient is low. The energy-saving objective is accomplished by including in the cost function the additional penalty term on the motor-to-battery regenerative braking power, while the safety objective is formulated as hard constraints on the longitudinal slip ratios of the wheels. Since the problem is based on a nonlinear vehicle longitudinal model, the global minimum within each time step is searched for by gridding the initial torque plane. Simulation results, based on a vehicle model in CarSim®, show that the proposed nonlinear model predictive controller is capable of restoring considerably more regenerative braking energy than a conventional proportional–integral controller supplemented with a feedforward control effort and another nonlinear model predictive controller with no consideration of the energy recovery and of maintaining a good vehicle-speed-tracking performance.