Optimizing Vehicle Approach Strategies for Connected Signalized Intersections
Optimizing Vehicle Approach Strategies for Connected Signalized Intersections
复制标题
优化互联信号交叉口的车辆接近策略
DOI：
10.1109/vnc48660.2019.9062810
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
D. Bondarenko
中科院分区：
文献类型：
--
作者：
Hendrik;Vivek Vijaya Kumar;Shahriyar Hussain;K. Sommerwerk;D. Bondarenko
在完全或部分自动驾驶车辆的背景下，接近交通灯管制的交叉路口或路段是一个挑战。需要可靠地检测即将到来的交通灯的当前状态，并且因此所得到的车辆接近行为只能基于交通灯的当前状态。虽然用于交通灯的基于摄像头的检测系统仅允许捕获交通灯的当前状态，但其剩余持续时间通常是未知的，就像手动操作的车辆一样。车联网（V2X）通信使车辆还能够直接从基础设施接收关于当前状态的剩余持续时间的信息，从而优化接近行为，以在当前或即将到来的绿色窗口内到达。虽然大多数相关的贡献旨在计算一个绿色光最佳咨询速度或提出一个计算昂贵的优化技术，这方面的贡献介绍了一个轻量级的和可扩展的方法，使连接（部分）自动化车辆的优化方法的产生。所引入的凸解空间概念的目标是低计算开销，并且仍然允许使车辆速度适应周围的交通参与者。
Approaching a traffic light regulated intersection or road segment represents a challenge in the context of fully or partially automated vehicles. The current state of an upcoming traffic light needs to be detected reliably and the resulting vehicle approach behavior can therefore only be based on the current state of the traffic light. While camera-based detection systems for traffic lights only allow for capturing the current state of the traffic light, its remaining duration is usually unknown, just as it is for manually operated vehicles. Vehicle-to-everything (V2X) communications enables vehicles to also receive information about the remaining duration of the current state directly from the infrastructure and to therefore optimize the approach behavior to arrive within the current or upcoming green windows. While the majority of related contributions aim at computing a green light optimal advisory speed or propose a computationally expensive optimization technique, this contribution introduces a lightweight and scalable methodology enabling the generation of an optimized approach behavior for connected (partially) automated vehicles. The introduced concept of a convex solution space targets a low computational overhead and still allows for adapting the vehicle speed to surrounding traffic participants.