A bi-level optimization framework for consecutive intersections under mixed traffic conditions based on corridor arrival-departure model

Autonomous Intersection Management (AIM) has attracted increasing research attention with the rapid advancement of Connected and Automated Vehicles (CAVs), Vehicle-to-Infrastructure (V2I) communication, and Internet of Things (IoT) technologies. Existing research demonstrates the potential for jointly optimizing signal schemes and vehicle trajectories to improve traffic efficiency at isolated urban intersections, while coordination of signal schemes and CAV trajectories among multiple intersections remains underexplored. Building upon our previous work on joint optimization at an isolated intersection, this paper proposes a bi-level optimization framework for urban corridors under mixed traffic conditions. At the upper level, a corridor arrival-departure model is developed to capture the dynamic relationships between consecutive intersections and to estimate total delay. The lower-level model, adapted from our previous work, takes the cycle lengths, phase orders, and reference green time durations from the upper level as input, and optimizes the final signal schemes and CAV trajectories at each intersection. A heuristic solution algorithm based on block coordinate descent is proposed to solve the upper-level problem. Simulations under varying traffic demands and CAV penetration rates are conducted, and the results demonstrate that the proposed method significantly enhances the traffic efficiency of urban corridors.