Intervehicle Force-Based Distributed Control of Merging Connected and Automated Vehicles at On-Ramps on Highways

Existing studies apply rule- or optimization-based approaches to determine the passing sequence at traffic merging points. However, the rule-based approaches [e.g., first-come-first-service (FCFS)] cannot guarantee the system optimality, and the optimization-based approaches can hardly handle heavy traffic due to the computational burden. This study proposes an “intervehicle force” based rule for connected and automated vehicles (CAVs) to cooperatively merge at one-lane on-ramps to balance the system optimality and computational efficiency. The control zone is divided into adjusting and stable zones. Vehicle passing sequence at a merging point is determined in the adjusting zone, and CAVs are mapped into one virtual lane in the stable zone according to the sequence. The “intervehicle force” model is formulated to determine a CAV’s acceleration rate profile based on its preceding CAV. Numerical experiments show that the proposed strategy significantly outperforms the FCFS strategy in terms of both vehicle throughput and average vehicle delay.