城市购物出发时间的非参数生存分析  

Nonparametric survival analysis of departure time of urban shopping trips

作　　者：李明[1] 杨小宝[1] 环梅[1] 董苓 

机构地区：[1]北京交通大学交通运输学院,北京100044

出　　处：《山东科学》2014年第4期85-91,共7页Shandong Science

基　　金：国家重点基础研究发展计划(973计划)(2012CB725400);国家自然科学基金(70901005;71071016;71131001)

摘　　要：根据济南市2011年的出行数据,研究了该市城市购物出发时间的分布特性.采用生存分析方法,建立城市购物出发时间的持续时间模型.运用非参数方法,对购物出发的持续时间进行估计,讨论了城市购物出行的时间分布和家庭社会经济属性变量对购物出发时间的影响.结果表明,总体样本的5.4％在7：00之前出发购物,69.7％在7：00～10：00出发购物,在10：00之后出发购物的少于24.9％.性别、年龄和家庭中儿童的数量对购物出发时间的选择有着显著的影响.女性比男性购物出发早,老年人更易在非高峰期进行购物出行,没有儿童的家庭购物出行在时间上更加灵活.本研究为购物出行的定量研究提供了准确、有效的分析工具,也对通过动态价格机制控制交通拥堵有重要意义.We addressed distribution characteristic of departure time of urban shopping trips based on the travel data of Jinan in 2011.We established its duration model with survival analysis method.We also estimated the duration time of shopping trips with a nonparametric method.We further discussed the impacts of departure time distribution of shopping trips and individual social-demographic properties on departure time of urban shopping trips.Results show that 5.4％ of the total samples depart for shopping before 7：00,69.7％ between 7：00 and 10：00,and less than 24.9％ after 10：00.Gender,age and children amount of a family have significant impacts on departure time：females are early than males in departure time; elders are easier to go shopping at non-peak hours; departure time is more flexible for a no-child family.Our results can provide an accurate and effective analysis tool for quantitative investigation of departure time of shopping trips and have guiding significance for the control of traffic congestion through dynamic pricing mechanism.

关 键 词：城市购物出发时间 连续时间模型 非参数生存分析 

分 类 号：U491[交通运输工程—交通运输规划与管理]