利用出租车轨迹数据估计城市道路拥堵状况

Estimation of urban traffic congestion by utilizing taxi trajectory data

交通拥堵是当代各大城市普遍面临的问题,如何通过各种技术手段发现城市交通的易拥堵路段并予以改善是城市交通相关部门以及学者关心的一个问题。通过各种GPS设备产生的历史轨迹数据对城市的研究逐渐受到重视,且取得一定的进展。文中总结运行车辆的4种通行状态:正常行驶、停靠/停泊、信号灯等待以及拥堵,分析这4种状态下运行车辆的速度特征,定义轨迹段平均速度和轨迹段速度方差两个数字特征用来描述运行车辆轨迹段的速度特征以及描述路段拥堵程度的拥堵指数,提出一种基于轨迹分段的拥堵检测的可视化方法,通过实验验证发现其可以较好地估计城市交通拥堵状况。

Traffic congestion is a serious urban problem commonly faced by the contemporary cities and how to find it by various technical means is a concern of urban traffic departments and scholars to improve the urban traffic status.Researches on urban problem by traj ectories data generated by various GPS devices draw more attention,and some progress has been made.This paper summarizes four running status for running vehicles：normal driving,docking/parking,signal waiting and congestion and then analyzes the features of vehicle speed of different running status by defining two numerical characteristics of speed and index of congestion.Finally,a visualization method for detecting traffic congestion is proposed,which proves to be effective by experiment.