School of Intelligent Engineering and Automation, Beijing University of Posts and Telecommunications, Beijing, China.
School of Systems Science, Beijing Jiaotong University, Beijing, China.
Accid Anal Prev. 2024 Dec;208:107798. doi: 10.1016/j.aap.2024.107798. Epub 2024 Sep 30.
Bottlenecks of the freeway generated especially by traffic accidents or temporary work zones contribute to significant reductions in system throughput and hinder the efficient traffic operations. It is imperative to take proactive measures to improve traffic state. With the rapid advancements in intelligent transportation, connected and autonomous vehicles (CAVs) have attracted much attention by its speculated capabilities in improving traffic safety and well-organized operational coordination. Therefore, reasonably utilizing the advantages of CAVs is possible to reduce the impact induced by bottlenecks. In this research, we propose a novel algorithm called CAV-Lead to obtain the CAV's regulated speed under mixed CAVs and human-driven vehicles (HVs) environment to improve the overall utilization of the freeway capacity near bottlenecks. Firstly, we illustrate the basic principle of the CAV-Lead algorithm that takes both microscopic and macroscopic traffic characteristics into account. Then, based on the local spatiotemporal traffic state, the CAV-Lead algorithm is proposed to determine each CAV's speed under mixed flow. Furthermore, a real-time simulation control framework considering the random behavior of HVs is presented. Moreover, several simulation evaluations including comparisons with basic scenarios and similar research are conducted under various CAV market penetration rates (MPRs). The results demonstrate that the CAV-Lead could improve the traffic performance, especially for the high traffic demand with certain MPRs.
高速公路的瓶颈，特别是由交通事故或临时工作区引起的瓶颈，会导致系统吞吐量显著减少，并阻碍交通的高效运行。采取积极措施改善交通状况至关重要。随着智能交通的快速发展，联网和自动驾驶车辆（CAV）因其在提高交通安全和有序运行协调方面的预期能力而受到广泛关注。因此，合理利用 CAV 的优势有可能减轻瓶颈造成的影响。在这项研究中，我们提出了一种名为 CAV-Lead 的新算法，以在混合 CAV 和人类驾驶车辆（HV）环境下获得 CAV 的调节速度，从而提高瓶颈附近高速公路的整体容量利用率。首先，我们说明了 CAV-Lead 算法的基本原理，该算法考虑了微观和宏观交通特性。然后，基于局部时空交通状态，提出了 CAV-Lead 算法来确定混合交通流中每辆 CAV 的速度。此外，还提出了一个考虑 HV 随机行为的实时仿真控制框架。此外，在不同的 CAV 市场渗透率（MPR）下，进行了包括与基本场景和类似研究进行比较的几项仿真评估。结果表明，CAV-Lead 可以改善交通性能，特别是在具有一定 MPR 的高交通需求下。
