Energy Management of Parallel Hybrid Electric Vehicles Based on Stochastic Model Predictive Control

This paper proposes a control approach for the energy management of parallel hybrid electric vehicles based on stochastic model predictive control (SMPC). Apart from minimizing fuel consumption, the controller additionally accounts for CO2 emissions. Considering the vehicle’s velocity to be time-varying, the limits for both propulsion machines of the hybrid vehicle are determined over a multiple prediction horizon. The stochastic approach has the advantage that the future driving profile does not have to be known in advance but is predicted based on an underlying stochastic model of the driver behavior. Simulation results obtained on standard driving cycles such as NEDC demonstrate the potential of the SMPC approach compared to a MPC controller with a-priori knowledge of the driving cycle.