Optimal Vehicle Following Strategy for Joint Velocity and Energy Management Control of Series Hybrid Electric Vehicles

Abstract

Recent advances in information and communication technologies present opportunities to optimally control the driving speed and powertrain energy management of vehicles under dynamic traffic circumstances. This paper addresses the energy-efficient car following problem of a series hybrid electric vehicle (HEV) by an enhanced adaptive cruise control (EACC) method. EACC is based on a nonlinear model predictive control framework, in which the behaviour of the lead vehicle is forecast by a neural network predictor trained by common test cycles. With the real-time predicted reference speed, EACC simultaneously optimizes the velocity and energy source power split of the ego HEV, while keeping the inter-vehicular distance within the desired range. The performance of EACC is benchmarked against a practical adaptive cruise control (ACC) that performs drafting and an impractical optimal control (OC) solved throughout the entire journey. Numerical examples show that the EACC can effectively close the gap between ACC and OC in terms of optimality with a remarkable fuel saving over ACC, while the computational load of EACC is comparable to ACC, which is much more efficient than the OC. Further design insight of the methodology is also provided by an investigation into the influence of the prediction horizon.