网联自动驾驶环境下交叉口车辆轨迹优化策略

Optimization Strategy of Vehicle Trajectory at Intersection in Connected and Autonomous Vehicle Environment

摘要:

网联自动驾驶技术为提升交叉口通行能力提供了新思路。针对信号交叉口，提出一种面向网联自动驾驶车辆的轨迹引导策略，以提高通行效率并降低行驶成本。首先，基于车辆路径及其潜在跟驰与超车成本，建立0-1整数规划模型，将交通流分配为多组车队。其次，分别构建纵向与横向轨迹控制模型：纵向模型综合考虑延误与燃油消耗；横向模型基于主从博弈理论，以实现车辆协同编队。最后，通过SUMO仿真软件验证策略有效性。结果表明，在交通量为每小时每车道350辆时，本策略相比仅速度引导策略，可使行驶时间与燃油消耗分别降低34.3%和30.1%，停车时间与停车次数分别减少38.7%和81.9%。

Abstract:

Connected and automated vehicle technologies offer new opportunities to improve the capacity of signalized intersections. This paper proposes a trajectory guidance strategy for CAVs at signalized intersections to enhance traffic efficiency and reduce travel costs. Firstly， a 0-1 integer programming model is established based on vehicle paths and potential car-following and overtaking costs， allocating the traffic flow into multiple platoons. Secondly， longitudinal and lateral trajectory control models are developed： the longitudinal model incorporates both delay and fuel consumption， while the lateral model employs a Stackelberg game theory approach to achieve cooperative vehicle platooning. Finally， the effectiveness of the proposed strategy is validated through simulations in SUMO. The results show that under a traffic volume of 350 per hour per lane， the proposed strategy reduces travel time and fuel consumption by 34.3% and 30.1%， respectively， and decreases stop time and number of stops by 38.7% and 81.9%， compared to a speed-only guidance strategy.