基于多源数据京津冀城市群边界识别研究

发布时间：2018-04-26 10:27

本文选题：城市群边界识别 + 大数据　；参考：《中国地质大学(北京)》2017年硕士论文

【摘要】：工业化以及城镇化进程的加快,将城市群变为区域发展的新单元,由此城市群的协调发展也成为了国家战略中重要的一项。精确地界定城市群的的边界,能够更加精确的规划和引导城市群的发展,包括城市群内部的城市化水平以及各个城市之间的分工体系,也包括城市群边界外围城市合理进入城市群等一系列工作。通信技术的迅猛发展带动了智能手机的大范围普及,从而将人类带入到大数据时代,从根本上改变了人们的生活习惯与日常活动,手机移动应用成为居民日常生活不可或缺的一部分,新浪微博作为社交类热门手机应用之一,具有较高的用户群体以及海量用户生成数据。由此可见,信息时代背景下,城市群边界的研究依然是城市群相关研究的重中之重,但是研究方法及数据源都存在新的探索。本文综述了城市群空间范围划定和边界识别的国内外研究现状,以及相关领域最新的研究动向,总结了现有相关研究中使用较多的城市群边界识别方法及其适用范围与弊端,说明了城市群边界在城市群研究乃至城市体系的相关研究中具有重要的研究意义。以此为基础,首先基于统计数据构建一套指标体系,并以京津冀城市群为例识别其边界,所识别的边界中共包括101个区县。然后采用新浪微博以及夜间灯光数据所构成的时空大数据,挖掘数据不同角度所携带的不同信息,构建一套由时空大数据不同角度的指标所组成的指标体系用于识别城市群的边界,其中包括城市联系强度、区县活动强度、区县集聚程度、区县灯光强度这4项指标,并以京津冀城市群为例实现了基于时空大数据识别城市群边界方法的探索,识别的边界包括88个区县。基于统计数据与时空大数据所计算得到的城市群值进行相关性分析,其结果呈现显著相关,这也证明基于时空大数据识别城市群边界的方法是客观可行的。通过对比两种数据源的结果,发现采用统计年鉴以及交通数据的结果更加倾向于经济要素,经济水平发展越高的区域,越有可能进入城市群内;而采用时空大数据识别的结果更加倾向于人口的分布情况,居民活动越频繁的区域,越有可能进入城市群内。
[Abstract]:The accelerated process of industrialization and urbanization has transformed the urban agglomeration into a new unit of regional development, and the coordinated development of the urban agglomeration has also become an important part of the national strategy. The precise definition of the boundary of the urban agglomeration can more accurately plan and guide the development of the urban agglomeration, including the level of urbanization within the urban agglomeration and the various types of urban agglomeration. The system of division of labor between cities also includes a series of work, such as the rational entry of urban agglomeration on the periphery of the border of urban agglomeration. The rapid development of communication technology has led to the widespread popularization of smart phones, thus bringing people into the era of big data, fundamentally changing people's living habits and daily activities, mobile mobile applications become residents. As one of the indispensable parts of daily life, Sina micro-blog, as one of the social popular mobile phone applications, has a high user population and massive user generated data. Thus, in the context of the information age, the study of urban agglomeration borders is still the most important of urban agglomeration research, but there are new research methods and data sources. In this paper, the research status of urban agglomeration and boundary recognition at home and abroad, and the latest research trend in the related fields are summarized, and the method and its application scope and malpractice of the urban swarm boundary identification used in the existing related research are summarized, and the correlation between the urban agglomeration and the urban system is explained. On the basis of this, we first set up a set of index system based on statistical data, and identify its boundary with Beijing Tianjin Hebei city group as an example. The identified boundary includes 101 districts and counties. Then, the space-time data made up of sina micro-blog and night light data can be used to excavate the different angles of the data. With different information, an index system composed of different angles of space-time data and different angles is constructed to identify the boundary of urban agglomeration, including the intensity of urban contact, the intensity of district and county activities, the degree of district and county concentration, and the light intensity of district and county, and the city group of Beijing, Tianjin and Hebei is used as an example to recognize the city based on space-time large data. The boundary of the group boundary method is explored, the boundary of the recognition includes 88 districts and counties. Based on the correlation analysis between the statistical data and the city group values calculated by the space-time large data, the results show significant correlation. It also proves that the method of identifying the urban group boundary based on the space-time large data is objectively feasible. By comparing the results of the two data sources, the results are found to be found. The results of statistical yearbook and traffic data tend to be more inclined to economic factors. The higher the economic level is, the more likely it is to enter the urban agglomeration, and the result of large data recognition is more inclined to the population distribution, the more frequent the residents' activities are, the more likely it is to enter the urban agglomeration.

【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F299.27;P208

【相似文献】