Ecological Predictive Cruise Control of Connected Electric Vehicle with Predecessor Velocity Prediction and Road Grade Preview

This paper proposes a predictive adaptive cruise control (PACC) strategy for connected electric vehicle (EV) which aims to minimize vehicle energy consumption with safe car-following. First, a Long Short-Term Memory (LSTM) is designed to predict the velocity of preceding vehicle. Second, the safe car-following spacing is derived by using the theory of state invariant set. Finally, an integrated PACC controller based on nonlinear model predictive control (NMPC) is proposed to achieve energy-efficient and safe vehicle cruising control. Simulation results show that the proposed PACC strategy can improve vehicle energy efficiency while maintaining a safe distance with preceding vehicles.