基于GIS的人口空间统计分析研究与实现

【作者】 何学洲；

【导师】 宫辉力； 李小娟；

【作者基本信息】 首都师范大学 ， 地图学与地理信息系统， 2008， 硕士

【摘要】 人口数据蕴藏着大量的信息,掌握人口的空间分布、迁移的空间特征、空间关联结构等对社会发展和政府决策具有重要的现实意义。如何挖掘提取这些有用的信息是当前人口学研究的重要问题。地理信息技术具有强大的空间分析处理能力,其在几十年的发展中已经广泛的应用于社会生产、生活中的各个领域,并在决策分析中发挥着越来越重要的作用。设计开发出为某一行业应用服务的专题地理信息系统,已是信息社会发展下主要的趋势。本论文利用地理信息空间分析技术,结合空间统计学的分析方法,针对人口的空间自相关结构、人口重心迁移、人口标准离差椭圆、人口分布集中指数PCI以及人口空间可视化表达进行研究,对部分方法进行改进,并利用现有的区域数据进行应用分析。本论文在相关理论算法的基础上,以Visual Studio 2005为系统开发的主体框架,使用C#程序设计语言,结合ArcEngine9.2、ArcGIS Server ADF组件技术,进行人口模型系统的设计与功能模块的开发,构建了具有客户端分析功能与服务器端专题数据发布功能的人口模型系统,解决了数据管理与数据分析脱离、数据分析与专题数据发布脱离的业务流问题。同时基于COM的思想,将功能模块设计成DLL组件,使得功能实现可以脱离模型系统,运行在其它的应用程序上,不但体现了系统功能上的灵活性,而且为以后模型系统的扩展,提供了很好的重构机能,具有一定的实用性。运用系统功能模块对研究区进行应用分析。选取京津冀地区1990年和2000年人口密度数据,进行空间自相关分析;分别计算得到两年的Moran’s全局自相关指数分别为0.1288和0.2033,经过正态分布的显著性假设检验,表明其人口的空间分布在整体上具有较强的正相关性,即呈现集聚状态。根据统计年鉴1990年至2006年北京市人口数据,进行北京市人口空间分布与迁移情况的分析;分析表明17年间北京人口重心总体向海淀区迁移,这和近年来海淀区经济迅速发展,吸引大量外来人口有着密切的关系。通过标准离差椭圆分析得出北京市人口呈“东北-西南”方向分布,由于17年来外来人口不断涌入,并定居在昌平、顺义、大兴、通州等北京市周边区县,并呈集聚状,故人口离差椭圆有缩小的趋势。但随着远郊区县的开发发展以及人口的不断涌入,远郊区县人口会不断增多,故预测离差椭圆将再次呈现扩张趋势,其覆盖区域将会扩大,即人口分布将呈现“扩散”状态。更多还原

【Abstract】 The demographic data contains a large amount of information , such as the distribution of population, character of population migration and the spatial auto-correlated of population, which is important and significant for the social development and the decision-making of the government. Therefore, how to dig these information is very important for the current research in demology. (The traditional analysis method of demographic data has already shown its limitation,so the more convenient, more flexible and more efective in precision and new analysis approaches of statistics are needed adapting to the sustainable development of the society.) Geographic information system (GIS) is a new science, which has a mighty competence to deal and analyses spatial geo-information. The Spatial Analysis with GIS is applied extensively in various fields during these decades, while it plays an important role in the analysis and decision-making. It becomes a trend to design a sort of geographic information system for the usage and service in some scopes.This paper analyses the spatial auto-correlated of population, the transfer of the population barycenter, the standard deviational ellipse of the population, the PCI index of the population distribution, and the visibility of the population spatial. I make use of the space-analysis technology in the geographical information , combining with the analyzing method of the space statistics, and making some improvement of these methods.Basing on the theoretical algorithm, Author draw a main frame by using the Visual Studio 2005, employ the C# program-design language, ArcEngine9.2, and ArcGIS Server ADF to design the model system and functional module. In this paper author construct a model system provided with the client-analyzing function and server data- promulgate function. Through this paper I have solved some problems on the operation, including the deviation between data management and data analysis, the data analysis renegading from the special data issuance. Meanwhile, founding on the COM, I design DLL groupware using parts of the functional module. This DLL groupware can deviate from the root model system, and be used on other applications. As a result, it shows the agility and practicability in systemic function, which can be used in the extensive or reconstructive system.In this paper, choosing the population density data of Beijing, Tianjin and Shanxi in 1990 and 2000. Using the systemic-function module, calculate the Globe Moran’s index in 1990 and 2000 is 0.1288 and 0.2033 respectively, After normal distribution significant hypothesis testing, make know that the population distribution relation of plus correlation are very intimately in this regions. According to the population data from 1990 to 2006, I study the population distribution and population transfer and find that during the 17 years the population center of Beijing transferred to haidian section as a whole, because of the development of haidian section and swarm of the population. Accoding to the standard deviational ellipse, I get the conclusion that the population of Beijing distributes as "northeast to southwest", and the people from other cities settled down in changing, sunyi, daxing, and tongzhou. Therefore, the standard deviational ellipse has the trend of becoming shrunken. However, along with the swarming of the population, the people will inhabitant in further sections and towns. Then forecast that the standard deviational ellipse will show the expanding trend again, covering more area, and the population will present as diffusing condition.更多还原