Cloud-Based Eco-Driving Predictive Cruise Control for Connected and Automated Electric Vehicles

The vehicle-road-cloud integration system (VRICS) offers enhanced information acquisition and computational processing capabilities, significantly advancing the development of connected and automated vehicles (CAVs). Leveraging the VRICS, this study proposes a cloud-based predictive eco-driving strategy (CPES) to improve energy efficiency and reduce intersection traversal time for CAVs. By integrating real-time roadside sensing and cloud-based historical data, the CPES enables CAVs to globally plan optimal lanes and driving speeds with receding horizon optimization, while explicitly accounting for the dynamic behaviors of surrounding vehicle queues. The CPES begins by developing a prediction model to estimate the dynamic queue dissipation time at signalized intersections, based on the characterization of both lateral and longitudinal movements of surrounding vehicles. Subsequently, an improved Grey Wolf Optimization algorithm is used to determine the optimal driving lane and speed of CAV under dynamic traffic conditions, incorporating the predicted queue dissipation into the optimization process. Finally, the CPES is validated through stochastic simulations and real-world experiments. Compared with the existing advanced methods, the results show that CPES achieves an average energy saving of up to 5.79% and reduces the average travel time by 4.65%. Furthermore, empirical vehicle testing demonstrates that CPES possesses practical deployment capabilities, maintaining robust performance even under complex and dynamic traffic conditions.