基于实时路况的西安交通拥堵时空分布特征研究

发布时间：2018-03-16 11:55

本文选题：城市交通拥堵　切入点：时空分布特征　出处：《西北大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着快速城镇化进程的不断推进,城市人口密度持续提高,城市机动车保有量呈“井喷式”增长,而与城市交通密切相关的道路里程的增长却远远落后。截止到2015年底,近十年间西安市机动车保有量增长293.6%,城镇常住人口增长19.73%,旅游接待人次增长336.2%,而西安市的公路里程却只增长39.85%,交通供需矛盾日益尖锐,交通拥堵逐渐发展成城市发展过程中必须解决的城市病之一。为了能够有效解决城市交通拥堵问题,从城市级别探究交通拥堵的时空分布特征,找出拥堵产生的机理十分必要,而随着大数据技术的不断发展,交通领域的实时路况数据因其具有更新快、数据量大、实时性高等特点而逐渐被研究者用来研究交通拥堵问题。本文参考国内外研究经验,提出了利用时空自相关和热点聚类进行拥堵热点判别的城市交通拥堵研究方法,结合西安市交通拥堵现状,利用高德交通平台发布的实时路况数据,从时间上和空间上分析了研究区交通拥堵的分布特征,以期为缓解交通拥堵提供参考。首先,论文利用莫兰指数、趋势分析、皮尔逊相关系数等方法探究了研究数据在时间和空间上的相关性和平稳性,以便为时空聚类分析提供前提。然后,论文将研究数据分为工作日、双休日和节假日三类,利用时间上的对比分析法和空间上的热点分析法分别对其进行研究,得出了以下结论:(1)整体上研究区交通拥堵在空间上呈现出一圈两点的空间特征。拥堵最为严重的区域位于二环路这一圈以及钟楼和小寨这两点,时间上呈现出明显的早晚高峰期。工作日时间上每天呈“M”型分布,空间上呈“环状”和“心形”分布;双休日时间上早高峰有所推迟,并增加了一个午间次高峰,空间上呈“蘑菇”状分布;节假日交通拥堵时空特征与双休日相似,同时叠加了旅游景区、景点的影响。(2)商圈分布、限行政策、人口密度以及路网密度与结构对研究区交通拥堵的时空分布特征均产生不同程度的影响。类似于钟楼、小寨的一些规模大、区位条件好、功能齐全的大型商圈更容易产生拥堵;限行力度与交通拥堵呈负相关,力度越大,对拥堵的缓解作用越明显;人口密度越大的区域交通拥堵也相应的越严重,二者呈正相关关系;路网密度越大越不容易造成交通,“棋盘+放射”状路网更容易产生交通拥堵。
[Abstract]:With the rapid development of urbanization, the density of urban population continues to increase, the number of urban motor vehicles is increasing "blowout", but the increase of road mileage, which is closely related to urban traffic, is far behind. As of end of 2015, In the past ten years, the number of motor vehicles in Xi'an has increased by 293.6, the resident population in cities and towns has increased by 19.73 and the number of tourists received has increased by 336.2, while the mileage of roads in Xi'an has only increased by 39.85, and the contradiction between supply and demand for transportation has become increasingly acute. Traffic congestion has gradually developed into one of the urban diseases that must be solved in the process of urban development. In order to effectively solve the problem of urban traffic congestion, it is necessary to explore the space-time distribution characteristics of traffic congestion from the city level and find out the mechanism of congestion. With the continuous development of big data technology, real-time traffic data in traffic field has been gradually used by researchers to study traffic congestion problems because of its characteristics of fast updating, large amount of data and high real-time performance. In this paper, a method of urban traffic congestion identification based on spatio-temporal autocorrelation and hot spot clustering is proposed. According to the traffic congestion situation in Xi'an, the real-time traffic condition data released by Gaude traffic platform are used. The distribution characteristics of traffic congestion in the study area are analyzed in time and space in order to provide a reference for alleviating traffic congestion. Firstly, the paper uses Moran index, trend analysis, Pearson correlation coefficient and other methods explore the correlation and smoothness of the data in time and space, in order to provide the premise for time and space clustering analysis. Then, the paper divides the data into three categories: working days, double weekends and holidays. Using the comparative analysis method in time and the hot spot analysis method in space, The following conclusions are drawn: 1) overall traffic congestion in the study area shows a space characteristic of two points in a circle. The most congested area is located on the second Ring Road and at the two points of the bell tower and Xiaozhai. During the working day, there was a "M" distribution and a spatial "ring" and "heart" distribution, while the morning peak was delayed and a midday peak was increased. The spatial distribution of traffic jams is similar to that of weekends, and it also superimposes tourist attractions, the influence of scenic spots, and the distribution of business circles and the restriction policy. The density of population and the density and structure of road network have different influences on the spatial and temporal distribution of traffic congestion in the study area. Similar to the bell tower, some large commercial areas with large scale, good location and complete functions are more likely to produce congestion. The intensity of traffic restriction is negatively related to traffic congestion, and the greater the intensity is, the more obvious the mitigation of congestion is; the more the population density is, the more serious the traffic congestion is, and the positive correlation is between them. The denser the road network is, the less likely it is to cause traffic congestion, while the "chessboard" road network is more likely to cause traffic congestion.
【学位授予单位】：西北大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.265

【相似文献】