Evaluating the driving behavior of autonomous vehicles and human driver vehicles in mixed driver environments at a signalized intersection
Evaluating the driving behavior of autonomous vehicles and human driver vehicles in mixed driver environments at a signalized intersection
复制标题
DOI：
10.1117/12.2613057
复制
发表时间：
2022-04
期刊：
影响因子：
3.2
通讯作者：
Xinyi Yang;Salman Ahmad;Martha K. Nelson;Joshua Rogers;Yihao Ren;Ying Hung;Pan Lu-
中科院分区：
文献类型：
--
作者：
Xinyi Yang;Salman Ahmad;Martha K. Nelson;Joshua Rogers;Yihao Ren;Ying Hung;Pan Lu-
随着智慧城市的快速发展，人们对车辆自动化的兴趣持续增长。自动驾驶汽车越来越受到人们的欢迎，被认为是地面交通的未来。配备自适应巡航控制 (ACC) 或协作式自适应巡航控制 (CACC) 的自动驾驶车辆为智慧城市中的智能交通提供了多种可能性。然而，在自动驾驶汽车在未来几十年内完全渗透市场之前，传统车辆和自动驾驶汽车将不得不共享相同的道路系统，这会因人类驾驶员的不一致而导致冲突。本文使用微型模拟器 VISSIM 评估了配备 ACC/CACC 的自动驾驶车辆和传统车辆在信号交叉口混合驾驶员环境中的性能。在仿真中，由 ACC/CACC 和 Wiedemann 99 (W99) 模型控制的车辆分别代表自动驾驶车辆和人类驾驶员车辆的行为。针对这两种不同的驾驶环境，综合考察了四种不同的运输方式：全轻型汽车、全卡车、全摩托车和混合工况。此外，每个模型都应用了十个不同的种子数以避免重合。为了评估人类驾驶员和自动驾驶车辆的驾驶行为，本文将基于明尼苏达州的一个真实道路交叉口，比较每个模型在信号交通交叉口的总停车次数、平均速度和车辆延误。
With the rapid development of smart cities, interest in vehicle automation continues growing. Autonomous vehicles are becoming more and more popular among people and are considered to be the future of ground transportation. Autonomous vehicles, either with adaptive cruise control (ACC) or cooperative adaptive cruise control (CACC), provide many possibilities for smart transportation in a smart city. However, traditional vehicles and autonomous vehicles will have to share the same road systems until autonomous vehicles fully penetrate the market over the next few decades, which leads to conflicts because of the inconsistency of human drivers. In this paper, the performance of autonomous vehicles with ACC/CACC and traditional vehicles in mixed driver environments, at a signalized intersection, were evaluated using the micro-simulator VISSIM. In the simulation, the vehicles controlled by the ACC/CACC and Wiedemann 99 (W99) model represent the behavior of autonomous vehicles and human driver vehicles, respectively. For these two different driver environments, four different transport modes were comprehensively investigated: full light duty cars, full trucks, full motorcycles, and mixed conditions. In addition, ten different seed numbers were applied to each model to avoid coincidence. To evaluate the driving behavior of the human drivers and autonomous vehicles, this paper will compare the total number of stops, average velocity, and vehicle delay of each model at the signalized traffic intersection based on a real road intersection in Minnesota.