窄路密网及智能网联交通环境下的跟驰数据集

本数据集采集自雄安新区窄路密网典型道路环境，记录了车辆纵向跟驰过程中的微观运行行为。数据时间范围为2023年至2025年，时间精度为0.1秒（10Hz），空间范围为河北雄安新区容东片区6个典型路段与交叉口群，空间精度优于10厘米。数据通过多源传感器（激光雷达、毫米波雷达、GPS/IMU、视频检测）融合采集，经轨迹重构与卡尔曼滤波处理，生成包含跟驰车辆与前车的轨迹、速度、加速度、车身尺寸等信息的高精度轨迹数据。数据集采用CSV格式存储，共包含多个车辆跟驰数据文件，适用于交通行为建模、跟驰模型标定、自动驾驶风险预测及交通仿真研究。数据质量控制严格，定位误差小，数据完整性高，具有重要的科研与工程应用价值。This dataset is collected from typical road environments characterized by narrow roads and dense road networks in Xiong'an New Area, documenting the microscopic operational behaviors during longitudinal vehicle car-following processes. The data spans from 2023 to 2025, with a temporal precision of 0.1 seconds (10 Hz). Its spatial coverage includes 6 typical road sections and intersection clusters in the Rongdong area of Xiong'an New Area, Hebei Province, with a spatial precision better than 10 centimeters. The data is collected via fusion of multi-source sensors including LiDAR, millimeter-wave radar, GPS/IMU, and video detection. After trajectory reconstruction and Kalman filtering processing, high-precision trajectory data is generated, which contains information such as the trajectories, speeds, accelerations, and vehicle body dimensions of both following vehicles and leading vehicles. The dataset is stored in CSV format, consisting of multiple vehicle car-following data files. It is applicable to traffic behavior modeling, car-following model calibration, autonomous driving risk prediction, and traffic simulation research. With strict data quality control, small positioning errors, and high data integrity, this dataset holds significant scientific research and engineering application value.

搜集汇总

背景与挑战

背景概述

该数据集采集自河北雄安新区容东片区的窄路密网典型道路环境，提供了2023年至2025年期间车辆纵向跟驰行为的高精度轨迹数据，时间精度为0.1秒，空间精度优于10厘米。数据通过多源传感器融合采集与处理，包含跟驰车辆与前车的轨迹、速度、加速度等信息，以CSV格式存储，适用于交通行为建模、自动驾驶风险预测及微观交通仿真等研究领域。