基于离散选择模型的城市轨道交通站点慢行交通吸引特性研究

发布时间：2018-03-25 00:25

本文选题：城市轨道交通车站　切入点：慢行交通吸引　出处：《兰州交通大学》2014年硕士论文

【摘要】：城市轨道交通作为现代城市交通的干线交通，是一种重要的绿色交通方式，与步行、自行车、电动车三种慢行交通，一同构成了城市居民出行方式的重要组成部分。本文主要是研究乘客在慢行交通换乘轨道交通过程中对慢行交通方式的选择问题，通过采用数据调查并建立ML离散选择模型，可为轨道交通站点对慢行交通客流吸引量的预测做基础性工作，为城市轨道交通站点布置和设计提供基础性数据和参考依据。通过研究城市轨道交通站点对周围区域的慢行交通量的吸引问题，并应用于轨道交通量预测，可使慢行交通与城市轨道交通换乘更协调，使城市绿色交通出行更为便利。论文首先从我国城市交通拥堵、轨道交通以及绿色交通的发展等角度入手，对研究背景进行阐述，分析了研究慢行交通与城市轨道交通换乘衔接问题的重要性与必要性。通过介绍国内外相关的研究成果，并进行分析总结，得出两点结论：对衔接轨道交通站点的慢行交通方式特性的研究不够；对衔接轨道交通站点的慢行交通方式吸引量预测研究不足。介绍了离散选择模型的基本理论知识以及SP与RP调查方法的基础理论知识。并以武汉轨道交通车站王家墩东站为例，进行数据调查工作，在调查方案设计中分别设计了RP调查内容与SP调查内容。在确定调查区域范围时，提出了“扇形分段区域抽样法”这一调查区域选定方法，确定了所要调查的区域，使调查区域更具有随机性与代表性。在慢行出行方式选择中，基于电动车的出行特点，提出将其归类于慢行交通方式，与步行、自行车一起构成慢行交通方式选择项集合。利用SPSS软件对各因素与慢行交通方式选择相关性进行了数据分析。通过计算各种影响因素的卡方检验数据，确定离散变量为慢行交通工具拥有种类、住址距轨道交通站的距离、交通服务水平，连续变量为出行费用。最后，根据调查数据，联合RP数据与SP数据建立融合离散选择模型，即ML模型。通过transcad软件对ML模型进行标定，结果显示：平衡系数检验值为3.792，较为显著，离散变量中乘客慢行交通工具拥有种类各检验值为：无自行车与电动车的t检验值为1.274，只拥有自行车的t检验值为1.650，，只拥有电动车的t检验值为1.703，其它变量的检验值也都比较显著。由此可认为所建立的RP/SP数据融合离散选择模型具有很好的应用参考价值，并可作为城市轨道轨道站点慢行交通吸引客流预测研究的参考。
[Abstract]:Urban rail transit, as the main line of modern urban traffic, is an important green transportation mode, and three kinds of slow-moving traffic, such as walking, bicycle and electric vehicle, This paper mainly studies the choice of slow transit mode in the process of slow transit transfer rail transit, and establishes ML discrete choice model by adopting data survey. It can do basic work for the prediction of passenger attraction to slow traffic at rail transit stations. This paper provides basic data and reference basis for the layout and design of urban rail transit stations. By studying the attraction of urban rail transit stations to the slow traffic volume in the surrounding area, the paper applies it to the prediction of the traffic volume of urban rail transit. It can make slow transit and urban rail transit more harmonious, and make urban green transportation more convenient. Firstly, from the point of view of urban traffic congestion, rail transit and the development of green traffic in China, the paper expounds the background of the research. This paper analyzes the importance and necessity of studying the connection between slow transit and urban rail transit, introduces the related research results at home and abroad, and makes an analysis and summary. Two conclusions are drawn: the study on the characteristics of slow transit mode connecting rail transit stations is not enough; This paper introduces the basic theoretical knowledge of discrete selection model and the basic theoretical knowledge of SP and RP investigation methods, and takes Wangjiadun East Station of Wuhan Rail Transit Station as an example. The contents of RP investigation and SP investigation are designed in the design of investigation scheme. When determining the scope of investigation area, the method of "sector sectional sampling" is put forward to select the investigation area. The area to be investigated is determined, which makes the investigation area more random and representative. In the selection of slow trip mode, based on the travel characteristics of electric vehicle, it is proposed to classify it into slow traffic mode and walking mode. Bicycles together form a collection of slow traffic mode options. The correlation between each factor and the choice of slow traffic mode is analyzed by using SPSS software. By calculating the chi-square test data of various influencing factors, the discrete variables are determined to be the type of slow traffic vehicle. The distance between address and rail transit station, the level of traffic service, and the continuous variables are the travel costs. Finally, according to the survey data, combining RP data with SP data, a fusion discrete selection model is established. ML model. The ML model is calibrated by transcad software. The results show that the balance coefficient test value is 3.792, which is significant. Among the discrete variables, the test value of the type of passenger slow-moving vehicle is 1.274 for non-bicycle and electric vehicle, 1.650 for bicycle only, 1.703 for electric vehicle, and 1.703 for other variables. It can be concluded that the discrete selection model of RP/SP data fusion has good reference value. It can be used as a reference for the prediction of slow-moving traffic in urban rail stations.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U239.5;U491.12

【参考文献】