基于虚拟流场与MSD模型的道路拥堵缓解算法

A Congestion Mitigation Algorithm Based on Virtual Flow Field and Mass-Spring-Damper Model

基于车间通讯的协同驾驶技术,能够有效提升智能交通系统的道路安全和通行效率,其中,利用协同换道技术解决瓶颈路段的拥堵问题正逐渐成为当前的研究热点.因此,提出了一种基于虚拟流场与质量—弹簧—阻尼模型(VFF-MSD)的道路拥堵缓解算法,通过模仿流体加速通过瓶颈的特性并结合虚拟的质量—弹簧—阻尼模型(MSD)对车辆换道过程进行控制,提高瓶颈路段的通行效率.首先,建立瓶颈路段的虚拟道路流场模型并进行流体力学仿真分析,得到流体的纵、横向速度数据,并根据这些数据确定车辆在道路上的换道策略;其次,建立换道车辆在目标车道上的投影,并在车辆投影与目标车道的前、后车辆间建立虚拟MSD模型,选取合适的弹簧刚度和阻尼系数保证协同换道过程的安全性;最后,控制所有车辆以最大速度和适当的安全车距通过瓶颈路段.仿真结果表明,所提出的算法在多车辆场景下可以有效缓解瓶颈路段的拥堵情况,使平均行驶时间减少11.52 s,平均速度提升6.4 m/s,通行效率提升12.17%.

With the help of vehicle-to-vehicle communication, autonomous vehicles(AVs) can improve road safety and traffic efficiency through collaborative driving. Among them, solving the congestion problem of bottleneck sections by using collaborative lane change has become a research hotspot. Therefore, a method based on virtual flow field and mass-spring-damper(VFF-MSD) is proposed in this paper. By imitating the characteristics of fluid accelerating through the bottleneck and combining with virtual mass-spring-damper model(MSD), the lane changing process of AVs is controlled to improve the traffic efficiency of the bottleneck section. Specifically, the data of longitudinal velocity and lateral velocity of the fluid was obtained by performing fluid simulation analysis on an established bottleneck of virtual road flow field model. Based on these data, the AVs’ lane change strategies on the road were determined. Then we established the projection of lane changing vehicles on the target lane, and virtual MSD model was established between the projection and the AVs in the front and rear of the target lane. Appropriate spring stiffness and damp parameters were selected to ensure the safety of cooperative lane changing process. Finally, all AVs were controlled to pass through the bottleneck at maximum speed and appropriate safe distance. Simulation result shows that the proposed algorithm can effectively alleviate congestion in multi-vehicle scenarios, reducing the average driving time by 11.52 s, increasing the average speed by 6.4 m/s, and improving the traffic efficiency by 12.17%.