Multi-vehicle Cooperative Control in Mixed Traffic Environments with Jointly Deployed Dedicated Lanes

To address the challenges of optimal dedicated lane allocation in mixed traffic environments, this study proposed a collaborative control framework for shared dedicated lanes that accommodate Connected and Autonomous Vehicles (CAVs) and buses. By analyzing the operational distinctions between curbside and bay-type bus stops, a bus clearance distance model was developed considering stop-type characteristics, thereby enabling the adaptive conversion of existing bus-only lanes into CAV-bus shared dedicated lanes. This methodology integrated a geometric configuration analysis of heterogeneous bus stops with dynamic lane management strategies to optimize infrastructure utilization, while prioritizing public transit efficiency. In a jointly-deployed dedicated (JDD) lane, each CAV was treated as an intelligent agent. The problem of multi-CAV cooperative control optimization in a mixed-traffic environment was modeled as a multiagent Markov decision process. Building on this, a multi-CAV cooperative control method based on the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm was proposed. Simulations conducted on the SUMO platform across various scenarios validated the proposed JDD lane strategy and the multi-CAV cooperative control method. The simulation results demonstrate that the MADDPG algorithm significantly enhances the cooperative control performance of CAVs in a dedicated lane environment. Specifically, the algorithm shows high traffic efficiency and stability across different CAV penetration rates, optimizing car-following and lane-changing strategies to reduce traffic conflicts and delays, thereby improving the overall road throughput. The cooperative effects among CAVs become particularly pronounced in high-penetration scenarios (above 60%), further optimizing traffic flow efficiency. The average traffic efficiency in the JDD lane environment increases by 9.42% and 5.61% compared to scenarios without dedicated lanes and with bus-only lanes, respectively. Moreover, the travel speed in the dedicated lane improves by 26.24%, while ensuring the priority passage of buses.