A Ground-Truth Dataset for the Validation of Traffic Demand Estimation

The paper provides two complete sets of data that can be used in estimating and validating traffic demand in a real motorway network in the urban area of Duisburg, Germany. The first set contains real data collected for a field study. The second set is gathered from a simulation model of the same motorway network to verify the results of the field study. Each dataset contains five files; (A): local detector data collected by inductive loops, and (B): trajectories of vehicles which were reconstructed from floating car data (FCD), (C) two network elements files (D): section detector data recorded by automatic number-plate recognition devices (ANPR). All data is processed, analyzed and aggregated if necessary. The processed dataset is presented in tables in separate CSV files. Additionally, the road network, to which all traffic data refers, is extracted from OpenStreetMap (OSM) and documented in a json file.

本论文提供两套完整数据集，可用于德国杜伊斯堡城区真实高速公路网络的交通需求评估与验证。第一套数据集为实地研究采集的真实数据，第二套数据集则取自同一高速公路网络的仿真模型，用于验证实地研究的结果。每个数据集包含五个文件：(A)感应线圈采集的本地检测器数据；(B)由浮动车数据（Floating Car Data, FCD）重构得到的车辆行驶轨迹；(C)两份网络要素文件；(D)由自动车牌识别设备（Automatic Number-Plate Recognition, ANPR）记录的路段检测器数据。所有数据均在必要时经过处理、分析与聚合。经处理后的数据集以表格形式存储于独立的CSV文件中。此外，所有交通数据所依托的道路网络均从OpenStreetMap（OSM）中提取，并以JSON文件形式归档。