Nonlinear model predictive control for improving energy recovery for electric vehicles during regenerative braking

This paper presents a nonlinear model predictive control (NMPC) scheme and a case study for improving the regenerative braking (RB) energy recovery for electric vehicles (EV) with in-wheel motors. The first part deals with a braking torque split problem, that is, given a desired vehicle longitudinal velocity profile, design braking torques for front and rear wheels independently to increase the RB energy recovery. The second part provides a case study to see the effects of different vehicle velocity profiles, with the same initial and terminal velocities and desired travelling distance, on RB energy recovery. The controller developed in the first part employs a three degrees-of-freedom longitudinal vehicle dynamic model with explicit considerations on the experimentally-measured, motor-to-battery RB efficiency map. Simulation results show that the proposed NMPC is capable of restoring more RB energy than a conventional PI controller does. The case study clearly shows the great potential in planning a priori velocity trajectory that is optimal in terms of energy recovery for RB control of EVs with in-wheel motors.