基于建成环境和低频浮动车数据的路段行程时间估计

Travel Time Estimation Based on Built Environment and Low Frequency Floating Car Data

将城市道路周边建成环境的相关属性作为路段行程时间的解释变量,结合城市低频浮动车数据,在不需要速度等GPS信息的条件下研究建成环境属性因素对路段行程时间的影响。同时,给出一种新的路段行程时间分布估计方法,即利用路段车辆数的分布代替路段长度作为路段行程时间的分配比例系数,得到路段行程时间的分布情况。为验证所提方法的正确性,以辽宁省丹东市振兴区锦山大街为例进行分析,用极大似然估计法得到各类建成环境对行程时间的影响参数值,并对比研究路段在有、无建成环境影响下的行程时间。结果表明:道路周边的建成环境会在不同时段导致路段行程时间显著增加,学校的影响时间段主要在6:00-7:20,医院、诊所集中在7:00-8:00,交叉口造成的行程时间增量在研究范围内整体较为平均。通过似然比检验,验证了将建成环境变量作为路段行程时间影响因素的可靠性。

This paper takes the relevant attributes of the built environment around the urban road as the explanatory variable of the road travel time,studying the impact on the travel time combined with low-frequency floating car data without complex GPS information,such as speed.At the same time,a new estimation method of link travel time distribution is proposed,which uses the distribution of the number of vehicles in the link as the proportional coefficient of link travel time distribution instead of its length,obtaining the distribution of link travel time.To verify the correctness of the proposed method,this paper takes Jinshan street in Zhenxing District,Dandong City,Liaoning Province as the example to obtain the impact parameters of the various built environment on travel time with the maximum likelihood estimation method.The results show that the built environment around the road will lead to a significant increase in the travel time of the road section in different periods.The impact time of schools is mainly from6:00 to 7:20,while hospitals and clinics are mainly from 7:00 to 8:00,and the travel time increment caused by intersections is relatively average in the whole research scope.Finally,through the likelihood ratio test,the reliability of taking built environment variables as the influencing factors of travel time is verified.