Driving simulator dataset on human driven vehicles' car following behaviour in mixed traffic with automated vehicles

This research is based on data gathered in 2023 at Delft University of Technology. This dataset is from a driving simulator experiment, whose goal was to study human drivers' behaviour in mixed traffic, which has both human-driven vehicles (HDVs) and automated vehicles (AVs).<br>A driving simulator experiment was designed to collect data on the car-following behavior of HDVs in the different scenarios. The driving simulator used is located at the Transport &amp; Planning department of Delft University in the Netherlands. It operates using the SCANeR (v1.9) software by AV Simulation. It is a fixed base driving simulator equipped with a Fanatec steering wheel and pedals, a dashboard mock-up, and three 4 K high-resolution screens which approximately provide 180° vision.<br>The dataset consists of 47 drivers. We categorized them into three age categories. There were 16 younger (25 - 45) drivers (8 male, 8 female), 16 middle-aged (45 – 65) drivers (8 male, 8 female), and 15 older (70+) drivers (10 male, 5 female).<br>The route in the driving simulator consisted of 3 parts. Part 1: 3 motorway on-ramps (excluding an initial on-ramp), Part 2: 3 provincial road signalized intersections, and Part 3: a straight road section. A depiction of the route is attached in this dataset information. Each driver drove four scenarios, excluding an initial familiarization scenario. The four scenarios varied in terms of the appearance of the vehicle interacting with the human driver and its driving style.<br>More details about the experiment set-up itself can be found in our paper published: TO BE ADDED<br><br><strong>Attached files:</strong>Processed datasetReadMe fileData processing scripts<br><strong>File formats:</strong>Data /.csvData /.xlsxPython scripts /.pyJupyternotebooks /.ipynb

本研究基于2023年于代尔夫特理工大学（Delft University of Technology）采集的实验数据。本数据集源自一项驾驶模拟器实验，旨在探究混合交通场景下人类驾驶员的行为规律，此类场景同时涵盖人类驾驶车辆（human-driven vehicles, HDVs）与自动驾驶车辆（automated vehicles, AVs）。<br>本实验旨在采集不同场景下人类驾驶车辆的跟驰行为数据，所用驾驶模拟器坐落于荷兰代尔夫特理工大学交通与规划系（Transport & Planning department），依托AV Simulation公司开发的SCANeR（v1.9）软件运行。该设备为固定基座式驾驶模拟器，配备Fanatec方向盘与踏板装置、仪表盘模拟装置，以及三块4K高分辨率屏幕，可提供约180°的视场范围。<br>本数据集共包含47名受试驾驶员，按年龄划分为三个组别：青年组（25-45岁）共16人（男性8名、女性8名），中年组（45-65岁）共16人（男性8名、女性8名），老年组（70岁及以上）共15人（男性10名、女性5名）。<br>驾驶模拟器中的行驶路线分为三个区段：第一区段含3条高速公路匝道（不含初始匝道），第二区段含3个省级道路信号灯控交叉口，第三区段为直线路段。本数据集附带该行驶路线的示意图。每名受试驾驶员需完成4个正式测试场景，不含初始熟悉场景。4个场景的差异体现在与受试驾驶员交互的车辆类型及其驾驶风格上。<br>关于实验设置的更多细节，可参阅待发表的研究论文：TO BE ADDED<br><br><strong>附加文件：</strong>已处理数据集、ReadMe自述文件、数据处理脚本<br><strong>文件格式：</strong>数据文件为.csv、.xlsx格式；Python脚本为.py格式；Jupyter笔记本为.ipynb格式