大规模路网车流数据集

由东南大学基于cityflow仿真平台数据格式，构建32×32 路网结构与车流仿真数据，用于cityflow仿真平台。路网文件构建1024个交叉口和连接交叉口的道路，每个交叉口参数包括中心位置、宽度、连接道路名称、可通行车道、相位控制方式等，每条道路参数包括位置、车道定义、两端交叉口编号等。车流量被特殊设计，以在整个交通网络中形成随着时间变化的不同的内部强耦合子区域来模拟城市交通动态。各道路均有直行车流和转弯车流穿越路网，平均到达率为 120 辆/小时。子区域内的交通流是随机生成的，具体来说，每个车流的起始和结束交叉口是从整个路网中随机选择的，平均到达率是从 [90, 240]中均匀采样的。模拟时间设置为 1500 秒。This dataset is constructed by Southeast University based on the data format of the CityFlow simulation platform, including a 32×32 road network structure and traffic flow simulation data, which is tailored for the CityFlow simulation platform. The road network file contains 1024 intersections and the connecting roads. The parameters of each intersection include central position, width, name of connected roads, passable lanes, phase control mode, and others. The parameters of each road include position, lane definition, numbers of the two connected intersections, and others. The traffic flow is specially designed to form dynamically changing internally strongly coupled sub-regions across the entire traffic network, aiming to simulate urban traffic dynamics. All roads support through traffic and turning traffic passing through the road network, with an average arrival rate of 120 vehicles per hour. The traffic flow within each sub-region is randomly generated: specifically, the origin and destination intersections of each traffic stream are randomly selected from the entire road network, and the average arrival rate is uniformly sampled from the interval [90, 240]. The simulation time is set to 1500 seconds.

搜集汇总

背景与挑战

背景概述

该数据集由东南大学基于cityflow仿真平台构建，包含一个32×32路网结构（含1024个交叉口）及车流仿真数据，用于模拟城市交通动态。车流设计考虑了随时间变化的内部强耦合子区域，平均到达率为120辆/小时，模拟时间为1500秒。