Integration of route updating and trajectory optimization for connected and automated vehicles on road networks

This study establishes a bi-level framework that integrates dynamic route planning and trajectory optimization for connected and automated vehicles (CAVs), aiming to reduce fuel consumption and travel delay. At the upper level, the optimal route for a CAV is updated upon entering a new link. Considering the differences in real-time traffic density and minimum travel time across lanes in different directions within the same link, a topological transformation approach is developed to convert the road network from a link-based to a lane-based structure. At the lower level, an optimization model is formulated to determine the ideal trajectory for a CAV within the optimization zone of a link. The actual trajectory is then adjusted from this ideal trajectory to meet safety requirements and traffic signal constraints. Moreover, the initial speed of a CAV upon entering the optimization zone is predicted and constrained in advance to prevent queue spillback within the optimization zone in case of congestion. The optimal route generated at the upper level serves as an input to the lower level for determining the vehicle’s terminal velocity relative to its direction, thereby forming one of the boundary constraints for the optimization model. Numerical simulations across four designed scenarios validate the effectiveness of the framework in improving energy efficiency and traffic performance for both gasoline and electric vehicles. The implementation of the proposed framework is expected to lead to reduced energy consumption and shorter travel times for vehicles in a fully CAV-equipped road network.