城市交通流量预测与拥堵治理数据集

采用交通流理论模型，基于路侧感知设备采集的红绿灯相位状态、车辆轨迹、速度、加速度等特征参数，实时计算流量、密度、速度、占用率等基础交通指标，通过加权平均法计算平均车速，统计车辆长度计算路段占用率。采用道路等级分类模型，根据道路类型（高速、主干道、次干道、支路）设定相应的饱和流量标准，结合车道数、绿信比参数计算路段通行能力。采用容量比模型，通过实际流量与通行能力的比值判定服务等级，划分为六个等级，分别对应畅通到严重拥堵的不同状态。采用时序预测模型，基于历史流量数据和当前交通状态，预测未来时段流量趋势，根据预测流量与通行能力的比值映射拥堵等级（畅通、缓行、拥堵、严重拥堵）。采用信号配时优化模型，基于实时服务等级和拥堵预测结果，动态调整信号周期、绿信比、相位相序等配时参数，实现信号配时的峰谷时段自适应切换，从而优化交叉口运行状态，提升通行能力和交通服务水平。Adopts traffic flow theory models, which calculate basic traffic indicators including flow, density, speed, occupancy rate and others in real time based on characteristic parameters collected by roadside perception equipment such as traffic light phase status, vehicle trajectories, speed and acceleration. The average vehicle speed is calculated using the weighted average method, and the road segment occupancy rate is calculated by counting vehicle lengths. Adopts road grade classification models, which set corresponding saturated flow standards according to road types (expressway, arterial road, secondary arterial road, branch road) and calculate the traffic capacity of road segments by combining parameters such as the number of lanes and green split. Adopts capacity ratio models, which determine service levels by the ratio of actual flow to traffic capacity, dividing them into six levels corresponding to different states from unobstructed traffic to severe congestion. Adopts time-series prediction models, which predict future traffic flow trends based on historical flow data and current traffic conditions, and map congestion levels (unobstructed, slow-moving, congested, severely congested) according to the ratio of predicted flow to traffic capacity. Adopts signal timing optimization models, which dynamically adjust timing parameters such as signal cycle, green split and phase sequence based on real-time service levels and congestion prediction results, realizing adaptive switching of signal timing during peak and off-peak hours to optimize intersection operating conditions and improve traffic capacity and service levels.

搜集汇总

背景与挑战

背景概述

该数据集基于天津（西青）国家级车联网先导区的路侧感知设备，汇集了各路口、路段的实时车流密度、速度等关键指标，构建高时空分辨率的交通流量数据集。通过交通流理论和机器学习模型，预测未来短时交通流量和拥堵程度，为交管部门提供精准的拥堵预警和信号配时优化建议，助力城市交通从被动疏导转向主动预防性治理。数据规模达近千亿条，每日更新，覆盖核心区域和主干路，适用于交通治理、调度和设施投资决策。