基于核查区的居民出行OD矩阵校核方法  

Verification method of resident trip OD matrix based on screen area

作　　者：樊大可[1] 罗小强[1] 陈宽民[1] 

机构地区：[1]长安大学公路学院,陕西西安710064

出　　处：《交通运输工程学报》2009年第6期121-126,共6页Journal of Traffic and Transportation Engineering

基　　金：国家863计划项目(2007AA11Z248)

摘　　要：为提高居民出行调查数据校核精度,按照组团状城市所布设的核查线可围合成若干封闭区域的特点,依据极大熵原理,提出了基于核查区的居民出行OD矩阵校核方法,量化分析了核查区进出交通量与OD矩阵分布量、核查区OD量与路段交通量之间的关系。以此为约束,按照核查线路段和核查区两个方面的观测值与估算值之间差异最小化为目标,建立了相应的模型及其算法,并应用2008年西安市居民出行调查数据对该方法进行了验证。计算结果表明:经校核后,519个交通小区的机动车出行OD矩阵分配量与观测交通量之间的平均相对误差为4.75%,误差最大值为12.28%,因此该方法可行。In order to improve the verification precision of resident trip survey, the characteristic of screen line was studied that could form a lot of closed areas in cluster city, and the verification method of resident trip OD matrix based on screen area was proposed according to maximum entropy principle. The relationship between the inner or outer traffic volume of screen area and OD matrix distribution, and the relationship between the OD matrix distribution of screan area and the traffic volume of road segments were analyzed. The difference minimum between the real values and estimation values in screen line segment and screen area was seen as target, and the corresponding model and algorithm were set up. The method was verified by the data of Xi＇an urban resident trip survey in 2008. Analysis result indicates that the mean relative error between the assignment volume of verified vehicle OD matrix and observed data is 4.75%, and the maximum error is 12. 28% in 519 traffic zones, so the method is valid. 8 tabs, 4 figs, 10 refs.

关 键 词：居民出行调查 数据校核 核查区 OD矩阵估计 

分 类 号：U491.1[交通运输工程—交通运输规划与管理]