基于球体格网的大规模三维矢量模型叠置分析方法研究

随着时空大数据的爆炸式增长，人们对于在全球尺度下分析理解时空大数据的需求也在不断增加。在利用高性能计算技术存储、处理、分析海量空间数据的同时，也应坚持发展地理信息系统的基本理论和算法。三维矢量模型叠置分析是一种重要的三维空间分析方法，其应用十分广泛；但由于三维空间中几何与拓扑的复杂性，其发展较为缓慢，复杂的三维矢量模型间的叠置易出错且效率低。本项目将针对大规模三维矢量模型叠置分析进行以下研究：（1）建立一种球体格网作为空间框架；（2）利用球体格网对全球海量三维矢量模型进行分层分块组织并建立索引；（3）在优化三维裁剪算法使其能更好地处理复杂三维矢量模型的基础上，研究三层过滤策略（包围盒、格网码、子块元数据）以减少不必要的裁剪运算，并结合分布式并行计算技术，提高大规模三维矢量模型叠置分析的效率。通过本项目的研究，将实现近实时的大规模三维矢量模型叠置分析，促进三维空间分析方法的发展。

With the explosive growth of spatiotemporal big data, people's demand for analyzing and understanding spatiotemporal big data on a global scale is also increasing. While using high-performance computing technology to store, process, and analyze massive spatial data, we should also adhere to the development of basic theories and algorithms of geographic information systems. Overlay analysis of three-dimensional (3D) vector models is an important 3D spatial analysis method, which has a wide range of applications. However, due to the complexity of geometry and topology in 3D space, its development is relatively slow, and the overlay of complex 3D vector models is prone to errors and low efficiency. In this project, the following research will be conducted on the overlay analysis of large-scale 3D vector models: (1) Establishing a sphere grid as a spatial frame. (2) Organizing global massive 3D vector models in layers and blocks with use of the sphere grid, and create the corresponding index. (3) Proposing an effective 3D clipping algorithm to better deal with complex 3D vector models, on which basis, the three-layer filtering strategy (bounding box, grid code, sub-block metadata) is going to be studied in reducing unnecessary clipping computing, and the distributed parallel computing technology will also be adopted to improve the efficiency of overlay analysis of large-scale 3D vector models. The research of this project will achieve the near real-time overlay analysis of large-scale 3D vector models, and thus promoting the development of 3D spatial analysis methods.