基于EAEQG格网模型的空间剖分组织方法研究
发布时间：2018-07-23 12:58
【摘要】：传统平面数据模型,是通过地图投影的方式将基于地理坐标系下的三维空间数据,转换为欧氏二维空间数据的数据模型。投影转换的方式决定了在大区域乃至全球尺度下,不可避免地会出现,空间数据的割裂、重叠和空间实体几何形变等现象。因而,传统平面数据模型难以实现对大区域的海量数据进行快速有效地存储、表达和管理。同时,随着对地观测手段的日益丰富,空间信息动态监测的区域范围已由局部扩展到了全球。空间数据,也渐渐成为多学科、多领域研究的数据基础。因此,如何快速有效地存储、表达和管理全球多尺度的空间数据,已成为了空间信息科学领域研究的重要问题之一。 以全球离散格网模型为理论基础的空间信息组织框架,被认为是解决传统平面格网模型在处理大区域乃至全球地理空间数据时的诸多问题最有效的方法。但是现行的全球离散格网三种剖分模型各自存在难以避免的问题： 1.自适应格网模型单一层面或单一比例尺的非规则剖分系统,不利于边界的动态操作,且缺乏数据的层次表达,很难与比例尺或空间分辨率直接联系,很难解决球面海量数据的多层次综合处理和空间分析； 2.正多面体格网模型由于最终要转换到椭球面上,因而涉及到复杂的投影运算,导致运算效率较低,难以保证数据高效转换。同时难以对现有数据进行集成和继承； 3.经纬度网格模型同一层级剖分单元的变形很大等诸多缺陷,难以适应现代信息技术日益多元化的应用需求。 因此,本文在总结现有全球离散格网模型优点和不足的基础上,创新性的提出了全新的格网剖分模型——等面积椭球面四尺树格网EAEQG(Equal-Area Ellipsoidal Quadtree Grid)模型。EAEQG格网模型是一种无缝无叠覆盖全球层次剖分格网,保证了同层次的剖分单元面积相等,并具有相同的几何形状和拓扑结构。实现了剖分单元和地理坐标快速转换,同时支持空间数据多分辨率表达。EAEQG剖分模型具体讲就是,选变经纬度剖分模型很好的继承了经纬度剖分模型的便于剖分单元编码与地理坐标间进行转换和对历史数据的良好继承性的优点。为了避免经纬度格网剖分模型会造成同一剖分层次下剖分单元变形较大的问题时,本文采用退化四叉树划分方法,控制同级剖分单元随纬度增加而变形增大的问题。同时,为了简化剖分计算的算法的复杂度,将正多面剖分思想引入EAEQG格网剖分。具体将就是,直接对WGS-84椭球体以内接八面体,进行初始剖分。即将旋转椭球体平均剖分成8个几何形态和特征完全相等的八分体。这样就将全球范围的剖分计算和操作算法,简化至对八分体内的剖分计算和操作算法,从而大大提高了操作和计算速率,系统的运行效率。采用四叉树划分,就决定了EAEQG剖分的剖分孔径为4,同时,也便于采用二进制的Morton编码,提高编码效率和剖分编码与地理坐标间互转换速率。EAEQG剖分基本剖分单元基本几何结构采用了规则的椭球面四边形,摒弃了三角形、六边形、菱形等规则几何结构,从而避免了正多面体剖分的剖分单元在投影映射到球面时的几何变形,最大限度的保证了各剖分单元的几何稳定性,以及避免了三角形、六边形、菱形等规则几何剖分单元难以与地理坐标进行转换的难题。同时,四边形简单拓扑关系使其更易于实现数据空间搜索和逻辑判断。最重要的是EAEQG格网剖分模型,在剖分的过程中,刻意地保留各剖分单元的等面积的几何特性。这一特性,使得EAEQG剖分数据在剖分计算和统计查询等方面具有无法比拟优势。 按照全球空间信息统一组织的思路,设计全新的空间信息剖分组织框架——基于EAEQG格网的空间信息剖分组织框架。并对EAEQG格网剖分的原理、结点坐标计算方法、剖分单元的编码方法、格网编码与地理坐标互转换算法、邻近搜索方法、几何特性分析、空间信息的剖分标识原理和原型系统的设计方案等方面进行了较为详细的阐述。 本文创新点为： ①WGS-84坐标系,作为空间参考基准。在WGS-84椭球面上直接进行剖分,很好的满足了现代空间信息处理对高精度的需求。 ②提出了全新的格网剖分模型EAEQG格网模型,弥补不现有全球离散格网模型的一些不足。
[Abstract]:The traditional plane data model is the transformation of three-dimensional spatial data based on the geographic coordinate system into the data model of the Euclidean two-dimensional spatial data through the way of map projection. The projection transformation determines the unavoidable appearance, the separation of space data, the overlap and the geometric deformation of space entities in large and global scales. As a result, the traditional plane data model is difficult to achieve fast and effective storage, expression and management of massive data in large areas. At the same time, with the increasing enrichment of ground observation means, the regional scope of dynamic monitoring of spatial information has been expanded from part to global. Spatial data has gradually become a multi-disciplinary and multi field research. Therefore, how to store, express and manage global multi-scale spatial data quickly and effectively has become one of the most important issues in the field of space information science.
The framework of spatial information organization based on the global discrete grid model is considered as the most effective method to solve many problems of the traditional plane grid model in dealing with large area and global geographic spatial data. However, the existing three kinds of discrete grid models are difficult to avoid.
1. the unitary or single scale irregular subdivision system of adaptive grid model is not conducive to the dynamic operation of the boundary, and it is very difficult to connect with the scale or spatial resolution directly. It is difficult to solve the multilevel comprehensive processing and spatial analysis of the massive data of the spherical surface.
The 2. positive polyhedron grid model has to be converted to the ellipsoidal surface, so it involves complex projection operations, which leads to low operational efficiency and difficult to ensure efficient data conversion, and it is difficult to integrate and inherit the existing data.
3. the longitude and latitude grid model has large deformation in the same level of subdivision unit. It is difficult to adapt to the increasingly diversified application demand of modern information technology.
Therefore, on the basis of summarizing the advantages and disadvantages of the existing global discrete grid model, this paper innovatively proposes a new grid subdivision model - the equal area ellipsoid four foot tree grid EAEQG (Equal-Area Ellipsoidal Quadtree Grid) model.EAEQG grid model is a seamless and non overlapping coverage global hierarchical grid, which guarantees the same The area of the level division unit is equal, and has the same geometry and topology. It realizes the rapid transformation of the division unit and the geographic coordinates, and supports the multi-resolution.EAEQG subdivision model of the spatial data, and the selection of the latitude and latitude subdivision model is a good inheritance of the subdivision unit coding. In order to avoid the problem of large deformation of the subdivision unit under the same section, the degenerate four fork tree division method is used to control the problem of the increasing deformation of the same level division unit with the increase of latitude. The complexity of the algorithm is simplified, and the idea of the positive multiplane section is introduced into the EAEQG grid. In particular, the initial dissection is carried out directly to the WGS-84 ellipsoid with eight sides, and the rotation ellipsoid is divided into 8 geometric shapes and the eight parts which are completely equal in feature. The calculation method is simplified to the calculation and operation of the division of eight points in the body, which greatly improves the operation and calculation rate and the efficiency of the system. By using the four forked tree division, the split aperture of the EAEQG section is determined to be 4. At the same time, the binary Morton coding is also used to improve the coding efficiency and the cross conversion between the split code and the geographical coordinates. The basic geometric structure of the basic dissection unit of the rate.EAEQG dissection uses a regular ellipsoid quadrangle, abandons the regular geometric structure of the triangle, hexagon, rhombus and so on, thus avoiding the geometric deformation of the subdivision unit when the projection is mapped to the sphere, and the maximum extent guarantees the geometric stability of the subdivision units. At the same time, the simple topological relation of the quadrangle makes it easier to realize data space search and logical judgment. The most important thing is the EAEQG grid partition model, which deliberately preserves the division units in the process of dissection. The geometric characteristics of the area make the EAEQG subdivision data incomparable in terms of subdivision calculation and statistical query.
According to the idea of global spatial information unified organization, we design a new spatial information dissection organization framework - the spatial information subdivision framework based on EAEQG grid, and the principle of EAEQG grid division, the calculation method of node coordinates, the coding method of the partition unit, the algorithm of grid coding and the mutual conversion of geographic coordinates, the adjacent search method, and the adjacent search methods The characteristic analysis, the partition principle of spatial information and the design scheme of the prototype system are described in detail.
The innovation points of this article are as follows:
(1) the WGS-84 coordinate system, as a spatial reference standard, is directly dissecting on the WGS-84 ellipsoid, which meets the high precision requirement of modern spatial information processing.
Secondly, we propose a new grid partition model EAEQG grid model to make up for some shortcomings of the existing discrete grid models.
【学位授予单位】：陕西师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208
本文编号：2139484
[Abstract]:The traditional plane data model is the transformation of three-dimensional spatial data based on the geographic coordinate system into the data model of the Euclidean two-dimensional spatial data through the way of map projection. The projection transformation determines the unavoidable appearance, the separation of space data, the overlap and the geometric deformation of space entities in large and global scales. As a result, the traditional plane data model is difficult to achieve fast and effective storage, expression and management of massive data in large areas. At the same time, with the increasing enrichment of ground observation means, the regional scope of dynamic monitoring of spatial information has been expanded from part to global. Spatial data has gradually become a multi-disciplinary and multi field research. Therefore, how to store, express and manage global multi-scale spatial data quickly and effectively has become one of the most important issues in the field of space information science.
The framework of spatial information organization based on the global discrete grid model is considered as the most effective method to solve many problems of the traditional plane grid model in dealing with large area and global geographic spatial data. However, the existing three kinds of discrete grid models are difficult to avoid.
1. the unitary or single scale irregular subdivision system of adaptive grid model is not conducive to the dynamic operation of the boundary, and it is very difficult to connect with the scale or spatial resolution directly. It is difficult to solve the multilevel comprehensive processing and spatial analysis of the massive data of the spherical surface.
The 2. positive polyhedron grid model has to be converted to the ellipsoidal surface, so it involves complex projection operations, which leads to low operational efficiency and difficult to ensure efficient data conversion, and it is difficult to integrate and inherit the existing data.
3. the longitude and latitude grid model has large deformation in the same level of subdivision unit. It is difficult to adapt to the increasingly diversified application demand of modern information technology.
Therefore, on the basis of summarizing the advantages and disadvantages of the existing global discrete grid model, this paper innovatively proposes a new grid subdivision model - the equal area ellipsoid four foot tree grid EAEQG (Equal-Area Ellipsoidal Quadtree Grid) model.EAEQG grid model is a seamless and non overlapping coverage global hierarchical grid, which guarantees the same The area of the level division unit is equal, and has the same geometry and topology. It realizes the rapid transformation of the division unit and the geographic coordinates, and supports the multi-resolution.EAEQG subdivision model of the spatial data, and the selection of the latitude and latitude subdivision model is a good inheritance of the subdivision unit coding. In order to avoid the problem of large deformation of the subdivision unit under the same section, the degenerate four fork tree division method is used to control the problem of the increasing deformation of the same level division unit with the increase of latitude. The complexity of the algorithm is simplified, and the idea of the positive multiplane section is introduced into the EAEQG grid. In particular, the initial dissection is carried out directly to the WGS-84 ellipsoid with eight sides, and the rotation ellipsoid is divided into 8 geometric shapes and the eight parts which are completely equal in feature. The calculation method is simplified to the calculation and operation of the division of eight points in the body, which greatly improves the operation and calculation rate and the efficiency of the system. By using the four forked tree division, the split aperture of the EAEQG section is determined to be 4. At the same time, the binary Morton coding is also used to improve the coding efficiency and the cross conversion between the split code and the geographical coordinates. The basic geometric structure of the basic dissection unit of the rate.EAEQG dissection uses a regular ellipsoid quadrangle, abandons the regular geometric structure of the triangle, hexagon, rhombus and so on, thus avoiding the geometric deformation of the subdivision unit when the projection is mapped to the sphere, and the maximum extent guarantees the geometric stability of the subdivision units. At the same time, the simple topological relation of the quadrangle makes it easier to realize data space search and logical judgment. The most important thing is the EAEQG grid partition model, which deliberately preserves the division units in the process of dissection. The geometric characteristics of the area make the EAEQG subdivision data incomparable in terms of subdivision calculation and statistical query.
According to the idea of global spatial information unified organization, we design a new spatial information dissection organization framework - the spatial information subdivision framework based on EAEQG grid, and the principle of EAEQG grid division, the calculation method of node coordinates, the coding method of the partition unit, the algorithm of grid coding and the mutual conversion of geographic coordinates, the adjacent search method, and the adjacent search methods The characteristic analysis, the partition principle of spatial information and the design scheme of the prototype system are described in detail.
The innovation points of this article are as follows:
(1) the WGS-84 coordinate system, as a spatial reference standard, is directly dissecting on the WGS-84 ellipsoid, which meets the high precision requirement of modern spatial information processing.
Secondly, we propose a new grid partition model EAEQG grid model to make up for some shortcomings of the existing discrete grid models.
【学位授予单位】：陕西师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208
【参考文献】
相关期刊论文 前6条
1 吴燕冈,申宁华,陈丽虹,范美宁;椭球面梯形面积的正反算问题研究[J];长春科技大学学报;1999年04期
2 林宗坚;关于构建数字地球基础框架的思考[J];测绘通报;1999年04期
3 明涛;庄大方;袁文;王占刚;;几种离散格网模型的几何稳定性分析[J];地球信息科学;2007年04期
4 李德仁,邵振峰,朱欣焰;论空间信息多级格网及其典型应用[J];武汉大学学报(信息科学版);2004年11期
5 郑彤;边少锋;;子午线弧长反问题新解[J];武汉大学学报(信息科学版);2007年03期
6 赵学胜,孙文彬,陈军;基于QTM的全球离散格网变形分布及收敛分析[J];中国矿业大学学报;2005年04期
，本文编号：2139484
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2139484.html
