基于动态时间弯曲距离的城市干道交通流量信息分析方法

Traffic volume information analysis for urban arteries using dynamic time warping distance

为研究城市干道间断流交通运行特征,采用动态时间弯曲距离作为相似性度量指标对城市干道交通流量信息进行分析。建立由数据预处理、动态时间弯曲距离计算、路径信息提取三步骤构成的干道流量信息相似性度量方法,应用实地观察数据,计算干道内不同检测点交通流量信息的相似性度量值,识别信息一致性良好的关键路径;比较同一路径内不同检测点交通流量信息的相似程度,检验信息的传递特征。该方法能自动区分城市干道系统内的关键路径与非关键路径,并反映交通流量信息传递过程中的衰减特征,用以确定干道信号协调控制的对象与范围。

To explore the traffic operating characteristic of urban arteries with several intersections,the dynamic time warping distance(DTW)was adopted as the similarity index to analyze the arterial traffic volume information.The similarity measurement was composed of three parts:data preprocessing,calculation of the dynamic time warping distance and extraction of the traffic stream information.With the field observation data,the critical traffic stream which has good consistent volume information was identified by calculating the similarity index of the traffic volume information from different detectors in the artery.The transfer characteristic was also testified by comparing the traffic volume information from different detectors of the same traffic stream.It proved that this method can distinguish the critical and non-critical traffic stream in the artery automatically,and depict the attenuation characteristic during the information transmission process,which helps to determine the object and scope of the arterial traffic cooperative control.5figs,11 refs.