摘要
以信号周期为时间窗的路段行程时间估计对交通运行状况分析具有重要意义。通过匹配路段上下游交叉口的自动车牌识别(ANPR,Automatic Number Plate Recognition)数据可以得到车辆的路段行程时间,使用缺失数据集获得的周期车均行程时间难以准确表征路段交通运行状况。因此本文提出一种基于周期的路段行程时间估计方法,该方法将匹配车辆的行程时间、到离上下游停止线的时刻、信号配时数据作为输入,建立基于最小二乘法的多段到达率行程时间模型,利用该模型对未匹配车辆行程时间进行估计。结果表明该方法能够较好地捕捉原数据特征,随着缺失车辆数的增多能够极大地减小周期车均行程时间误差,并且在79.99%的情况下有正收益,20.58%的情况下收益值大于10s。
Estimating travel time for links using signal cycles as time windows is of significant importance for analyzing traffic operations.By matching Automatic Number Plate Recognition(ANPR)data at intersections upstream and downstream of links,the travel time for vehicles on those links can be obtained.However,using average travel time based on incomplete data sets does not accurately represent the traffic conditions of links.Therefore,this paper proposes a cycle-based method for estimating travel time on links.This method utilizes matched travel time data,arrival and departure times at upstream and downstream stop lines,and signal timing data as inputs to establish a multi-segment arrival rate travel time model based on the least squares method.This model is then used to estimate travel time for unmatched vehicles.The results indicate that this approach effectively captures the intrinsic data characteristics,leading to a substantial reduction in the error of average travel time for links as the number of missing vehicles increases.Moreover,it exhibits positive gains in the case of 79.99%missing vehicles,while in the case of 20.58%missing vehicles,the gains surpass 10s.
作者
田玲玲
TIAN Lingling(School of Transportation and Logistics,Southwest Jiaotong University,Chengdu 610031,China)
出处
《综合运输》
2024年第6期116-122,共7页
China Transportation Review
关键词
ANPR数据
周期行程时间估计
最小二乘法
多段到达率
未匹配车辆
ANPR data
Cycle-based travel time estimation
Least Squares Method
Multi-segment arrival rate
Unmatched vehicles
