公交动态信息下通勤出发时间选择行为分析  

Analysis of the Choice Behavior of Commuting Departure Time under the Dynamic Transit Information

作　　者：郭士永[1,2] 李文权[1,2] 陈茜[1,2] 左毅刚[1,2] 

机构地区：[1]东南大学城市智能交通江苏省重点实验室,南京210096 [2]现代城市交通技术江苏高校协同创新中心,南京210096

出　　处：《交通信息与安全》2014年第3期12-17,共6页Journal of Transport Information and Safety

基　　金：国家重点基础研究发展(973)计划项目(批准号:2012CB725402);国家青年自然科学基金项目(批准号:51208099)资助

摘　　要：在交通行为分析中,传统效用函数形式不能反映个人属性对信息效用感知的影响。改进效用函数形式,运用多项Logit模型理论,建立了公交动态信息下的公交通勤者出发时间选择行为初始模型。从初始模型中选择t检验显著的解释变量,并分别基于传统效用函数形式、改进的效用函数建立了2个修正模型。对比2个修正模型标定结果发现,改进的效用函数比传统的效用函数更有利于提高模型性能。研究也表明:性别、收入的t检验值分别为2.02、-2.25,说明二者对乘坐时间信息效用感知显著;教育属性的t检验值为2.03,说明其对车内拥挤程度信息效用感知影响显著;乘坐时间取均值,车内拥挤度为0.7时,二者效用比值为1.54∶1,说明乘坐时间信息是影响出发时间选择最重要的因素。The traditional utility function form cannot be used to study the influence of individual attributes on the perceived utility of transit information .In order to overcome this issue ,a modified utility function form is developed to study the impact of dynamic transit information onto the choice behavior of commuting departure time .According to the results of the preliminary model ,independent variables whose t-test are significant are kept and used to develop two modi-fied models ,with one using traditional and the other using modified utility function form respectively .The comparison be-tween the two models shows that the one with the modified utility function form has a better model performance than the one with the traditional utility function form .The study results also indicate that the gender and income attributes ,whose t values are 2 .02 and-2 .25 ,have a significant impact on the perceived utility of riding time .In addition ,the education level has ,which has at value of 2.03,a significant impact on the perceived utility of crowding level .It is also found out that ,when the riding time takes its mean and the crowding level is 0.7 ,their utility ratio is 1.54∶1 .The ratio reveals that the riding time is the most important factor to influence the departure time choice of transit commuters .

关 键 词：公交通勤 出发时间选择行为 多项Logit模型 公交动态信息 

分 类 号：U491.1[交通运输工程—交通运输规划与管理]