Vehicle–Infrastructure Cooperative Multi-Lane Lane Changing of Connected and Automated Vehicle Platoons in Mixed Traffic via Monte Carlo Tree Search

Connected and automated vehicle (CAV) platooning can improve freeway traffic efficiency, energy performance, and driving stability, but these benefits depend on safe and efficient lane changing in mixed traffic. The problem becomes much more challenging when a platoon must traverse multiple lanes, because it requires sequential decisions on feasible gap selection and lane-changing order, while the platoon may be temporarily split and later reformed after intermediate transitions. To address this problem, this paper proposes a vehicle–infrastructure cooperative framework for platoon multi-lane lane changing in mixed traffic. At the infrastructure side, a centralized decision layer uses Monte Carlo tree search to determine the target gap assignment and lane-changing sequence based on traffic perception information. At the vehicle side, a distributed planning layer generates executable cooperative actions through a joint optimization model that coordinates the active lane-changing CAV and the matched yielding CAV under short-term human-driven vehicle motion prediction. In this way, the proposed framework links global maneuver organization with vehicle-level execution. Simulation results on a four-lane freeway show that the proposed method can successfully organize and execute progressive multi-lane maneuvers, during which the platoon may be temporarily split and then reformed in the target lane. Additional tests under varying traffic demand, platoon size, and emergency conditions further demonstrate its adaptability and robustness. Compared with a benchmark obtained by repeatedly applying a single-lane platoon lane-change strategy, the proposed method consistently reduces both maneuver time and required travel distance, with larger gains for larger platoons. The results demonstrate the effectiveness of explicitly addressing continuous multi-lane lane changing and highlight the value of vehicle–infrastructure cooperation for CAV platoons in mixed traffic.