摘要
为在城市智能公交建设经费有限的前提下科学的选择最有必要的公交站点投放建设电子站牌,本文提出一套基于公交大数据及GIS构建的电子站牌投放综合选择指标体系。在此基础上,提出主成分-Two Step聚类组合模型基于指标体系结果进行公交站点的分类、排序。应用2015年3月2日到2015年3月15日共计14 d的哈尔滨市公交大数据进行案例分析,提出在公交IC卡与GPS系统存在时间差且无对照匹配字段的前提下的大数据匹配算法。根据案例研究结果显示,在应用本文提出的方法后,哈尔滨市1 553个公交站点按照布设电子站牌需求程度由高到低分为了5类,第1类布设电子站牌优先级最高,包含了135个站点,但仅占总体的8. 69%;最后1类布设需求最低,包含了959个站点,占总体61. 75%。首先,通过优度和聚类结果距离等指标验证本文所构建的模型获得较好的效果;此外,经过与哈尔滨实际情况的对比分析,验证本文所提出的电子站牌投放综合选择模型合理、有效和可应用性强,能够很好的区分站点布设电子站牌需求程度的高低。该模型可为各城市经费有限情况下投入建设公交电子站牌提供科学、有效的方法。
In order to scientifically select the most necessary bus stops to put electronic bus station boards on the premise of limited funds for urban intelligent bus construction,this paper proposes a comprehensive selection index system for electronic bus station boards based on bus big data and GIS.On this basis,a principal component-TwoStep clustering combination model is proposed to classify and rank the bus stops.In this paper,the Harbin bus data of 14 days from March 2,2015 to March 15,2015 were used for case analysis.A data matching algorithm based on the time difference between IC card and GPS systems without public matching field was proposed.According to the case study results,1 553 bus stations in Harbin were divided into five categories according to the degree of demand for bus electronic station board.The first category had the highest priority,including 135 stations,but only 8.69% of the total;the last category had the lowest demand,including 959 stations,accounting for 61.75% of the total.Firstly,the model constructed in this paper was validated by the indexes of fit and distance of clustering results.In addition,by comparing and analyzing with the actual situation in Harbin,it was proved that the comprehensive selection model of electronic station board put forward in this paper was reasonable,effective and applicable,and can distinguish the level of demand for electronic station board well.The model can provide a scientific and effective method for the construction of bus electronic station board in the case of limited funds for each city.
作者
朱权
张荣川
段利敏
ZHU Quan;ZHANG Rongchuan;DUAN Limin(Kunming Urban Transport Institute,Kunming 650000;Dali Municipal Public Security Bureau Traffic Police Brigade,Dali 671000;Kunming Municipal Public Security Bureau Traffic Police Detachment,Kunming 650000)
出处
《森林工程》
2019年第4期89-96,共8页
Forest Engineering
基金
住房城乡建设部科学技术项目(2015-R2-014)
关键词
公交电子站牌
大数据
综合选择模型
聚类分析
Bus electronic station board
big data
comprehensive selection model
cluster analysis
