公交客流统计时段的划分方法  

Time Division Method in Statistics of the Volume of Bus Passengers

作　　者：范海雁[1] 姚佼[1] 

机构地区：[1]上海理工大学管理学院,上海200093

出　　处：《森林工程》2014年第6期141-143,共3页Forest Engineering

基　　金：同济大学道路与交通工程教育部重点实验室开放基金项目(K201303)

摘　　要：在城市公共交通运营研究中,公交客流数据的取得和统计工作是非常重要的,而确定高峰统计时段是最基础的工作。简述了公交发车间隔研究的重要性,在计算城市公交线路的发车间隔时间时,需要先计算统计时段的客流需求量。而统计时段的划分是文章重点研究的问题,基于聚类分析,建立一种均值聚类算法,得到具有相似客流特征的时间段,并以公交线路875路的跟车调查数据为基础,在上海市公交875路客流的实际统计中进行应用。通过实证研究,表明,可以有效减少聚类数据的冗余信息,特别适用解决大数据聚类问题,其结果可以自动划分客流时段,时段划分结果为0.5 h或1 h的非整点时段,对于统计时段内的高峰小时的确定和行车间隔的优化具有重要的意义。While studying the operation of city public transit, the data acquisition and statistical work of public transport passenger flow is very important and the determination of the peak period is the most fundamental work. In this paper, the importance of analyzing the bus interval time is introduced. The calculation of the passenger demand flow in statistical periods is the prerequisite for the optimization of interval time. A k-means clustering algorithm is established based on the cluster analysis, which has similar time peri- od of passenger flow characteristics, and the application is used in the actual statistics passenger flow of bus line 875 in Shanghai. The results show that it can effectively reduce the redundant information of the clustering data, especially suitable for the clustering problem of large data. It can automatically divide the flow periods into half-hour periods and hourly period, which is very important to determine the peak hours and optimize the vehicle interval time.

关 键 词：公共汽车 客流统计 时间划分 优化 

分 类 号：U492.2[交通运输工程—交通运输规划与管理] S782.5[交通运输工程—道路与铁道工程]