江海东, 张可能. 空间划分多因子耦合PDE模型与算法[J]. 武汉大学学报 ( 信息科学版), 2016, 41(8): 1060-1065. DOI: 10.13203/j.whugis20130747
引用本文: 江海东, 张可能. 空间划分多因子耦合PDE模型与算法[J]. 武汉大学学报 ( 信息科学版), 2016, 41(8): 1060-1065. DOI: 10.13203/j.whugis20130747
JIANG Haidong, ZHANG Keneng. Multi-factor Coupled PDE Model and Algorithm for Spatial Partition[J]. Geomatics and Information Science of Wuhan University, 2016, 41(8): 1060-1065. DOI: 10.13203/j.whugis20130747
Citation: JIANG Haidong, ZHANG Keneng. Multi-factor Coupled PDE Model and Algorithm for Spatial Partition[J]. Geomatics and Information Science of Wuhan University, 2016, 41(8): 1060-1065. DOI: 10.13203/j.whugis20130747

空间划分多因子耦合PDE模型与算法

Multi-factor Coupled PDE Model and Algorithm for Spatial Partition

  • 摘要: 为研究非凸空间离散数据的空间划分,建立了耦合数据集的凸凹性、数据规模、离散度的空间划分的数学模型。利用Laves划分标识36对非凸空间离散数据进行有限区域变比例划分,然后通过地形曲面微分单元与数据规模的偏导函数关系,耦合离散度计算空间划分的单元间距和数量。最后通过构建DEM,可视化验证和对比分析发现,耦合模型能够计算出非凸离散空间数据空间划分单元的间距和数量,也能实现不同分辨率的划分单元的无缝拼接;且当试验数据从110组递增至440组时,该模型耗时仅是44标识划分和Delaunay的1/10~1/3,且随数据规模成倍增加时耗时基本呈线性增长,收敛性较好,但耗时随离散度增加而增长。

     

    Abstract: In order to study spatial partition of the non-convex discrete spatial data, established mathematical model coupling the convexity, data size and dispersion of the dataset. Non-convex discrete spatial data was divided in the limited area by Laves divided identity 36, and then by the partial derivatives relationship of the terrain surface between the differential unit and the data density, the spacing and number of cells was calculated coupled discrete.Last visualization and compared analysis through building DEM showed that: coupled model can calculate out the spacing and number of non-convex discrete spatial data, also achieved seamless mosaic of divided units for different resolution; and while test data from 110 groups to 440 groups, the time cost of this paper was only 1/10-1/3 of 44 divided identifies and Delaunay, and was linear growth with the data multiplied increased, besides the convergence was better, however it took growth with discrete degrees increased.

     

/

返回文章
返回