路段平均行程时间估计方法  

Estimation method of average travel time for road sections

作　　者：张和生[1] 张毅[2] 胡东成[2] 

机构地区：[1]北京交通大学电气工程学院 [2]清华大学自动化系,北京100084

出　　处：《交通运输工程学报》2008年第1期89-96,共8页Journal of Traffic and Transportation Engineering

基　　金：国家863计划项目(2006AA11Z231);国家973计划项目(2006CB705506);国家自然科学基金项目(60374059);轨道交通控制与安全国家重点实验室开放课题项目(SKL2007K011);北京交通大学科技发展基金项目(DQJ05007)

摘　　要：为了有效利用线圈检测数据,精确估计路段平均行程时间,提出了一种路段平均行程时间估计方法。将路段平均行程时间分为平均行驶时间、平均排队时间和平均通过路口时间三部分。考虑线圈埋设的特点,通过估计平均行驶速度得到平均行驶时间。用分段时齐Poisson过程描述车辆驶入路段过程和驶离过程,用Markov排队模型描述车辆排队过程,用生灭过程描述排队车辆数,得到车辆排队模型,计算了路段有、无初始排队的平均排队时间。基于选取与路口相关的饱和流率和平均车长,计算了平均通过路口时间。计算结果表明:平均行程时间估计值与实测值的误差小于12%,说明路段平均行程时间估计方法可行。In order to availably utilize loop data and estimate the average travel time of road sections accurately and quickly, an estimation method of average travel time was proposed. Average travel time was divided into average running time, average queue delay and average crossing intersection time. Average running time was calculated by using average speed based on taking account of the installation type differences of loops. Piece-wise time homogeneous Poisson process was used to describe the arrival and departure processes of vehicles, Markov model was used to describe the queue process of vehicles, birth and death process was used to model vehicles＇ number in queue, and two kinds of average queue times with initial queue and no initial queue were computed. Average crossing intersection time was computed based on average vehicle length and saturation volume rate at intersection. Computation result shows that the relative error between the measure value and computation value of average travel time is less than 12 %, so the method is feasible. 1 tab, 9 figs, 14 refs.

关 键 词：智能交通系统 平均行程时间 线圈数据 Markov排队模型 POISSON过程 

分 类 号：U491.1[交通运输工程—交通运输规划与管理]