Synthetic 10-Hour Traffic Simulations for Central Athens (Base/Lane Closures/Rain)

This release provides 18 CSV trace files from 10-hour microscopic traffic simulations of central Athens, generated with the SUMO (Simulation of Urban MObility) toolkit. Three distinct scenarios capture different operating conditions: Base – the unmodified network under typical morning and afternoon peak volumes. Closure – the same base network with lane closures on key central streets to emulate construction or incidents. Rain – the base network under simulated rainfall, modeled via reduced free-flow speeds, increased following gaps and more. Each scenario is split into train and test runs: the train set uses one random seed for trip generation and baseline traffic-generation periods, while the test set uses a different seed and perturbs the per-period volumes by random noise to introduce realistic variability. For each of the three scenarios (base, closure, rain) and each split (train, test), you’ll find three CSV outputs: Dump - a full state dump, organized by timestep > edge > lane > vehicle (id, position, speed) FCD - floating car data, organized by timestep > vehicle (id, x, y, speed, acceleration, odometer) Emission - emissions trace, organized by timestep > vehicle (id, fuel, waiting time) The following SUMO tools were used: osmWebWizard - exporting the network from OpenStreetMap. netedit - selecting the road closures. randomTrips - generating the random trips based on traffic generation periods. duarouter - routing a few fixed routes that were added on each test set for later visualizations.本数据集发布了18个CSV格式的轨迹文件，均基于雅典市中心10小时微观交通仿真生成，仿真工具采用SUMO（Simulation of Urban MObility，城市移动性仿真工具包）。 本次仿真共设置三种差异化运行场景，覆盖不同交通工况： 1. 基准场景（Base）：未修改的原始路网，采用典型早、晚高峰时段的交通流量。 2. 道路封闭场景（Closure）：在基准路网的核心街道设置车道封闭措施，模拟道路施工或突发交通事件。 3. 降雨场景（Rain）：在基准路网基础上叠加仿真降雨工况，通过降低自由流速度、增大跟车间隙等方式实现交通行为建模。 每个场景均划分为训练集与测试集：训练集使用单一随机种子生成出行需求与基础交通流量时段；测试集则采用不同的随机种子，并为各时段的交通流量添加随机噪声以引入符合真实情况的交通波动。 针对三种场景（基准、道路封闭、降雨）以及两种数据集划分（训练集、测试集），共包含三类CSV输出文件： - 状态转储文件（Dump）：完整的系统状态记录，按时间步>路段>车道>车辆的层级结构组织，包含车辆ID、位置、速度信息。 - 浮动车数据文件（FCD，Floating Car Data）：按时间步>车辆的层级组织，包含车辆ID、x坐标、y坐标、速度、加速度、里程数信息。 - 排放轨迹文件（Emission）：排放相关轨迹数据，按时间步>车辆的层级组织，包含车辆ID、燃油消耗量、等待时长信息。 本次仿真采用以下SUMO官方工具完成： - osmWebWizard：从OpenStreetMap导出路网数据 - netedit：用于选定并设置道路封闭路段 - randomTrips：基于预设的交通流量时段生成随机出行需求 - duarouter：为各测试集生成若干固定路由，用于后续可视化工作