The Research And Realization On The Spatial Computing Models For Huge Spatial Data

In the present information age, in the field of intelligent information technology, information technology makes our lives convenient. At the same time, we also have found that information dealt in daily life is extremely large, including geospatial data which has location spatial information, it is also used in very large areas, spatial data toward the continued large and expansion direction development.With the development of geographic information systems, social information needs the growing prosperity, geographic information system software will deal with the increasingly large range of large-scale data. For massive data, the efficient storage, searches and analyse GIS spatial data has always been a hotindustry research and difficult problems, but is also the innovation problem competed to raise for the major GIS softwares.To further enhance the torage, searches and analysis mass spatial data's capabilities for GIS software, and to provide theoretical reserves and core technology solutions, this project aims base on spatial computing model, proposed a set model solution of GIS software to enhance storage, searches large-scale vector spatial data analysis and improve the efficiency of spatial analysis, without changing the core of spatial analysis algorithms, based on the model layer from the top:to treat analysis of spatial data from the data pre-structured form on the different operations of spatial analysis, and summarized optimal general conditioning solution, optimize the physical storage, index spatial data, and propose a grid framework strategies in parallel. Solve the spatial analysis efficiency for large-scale spatial data.Spatial analysis is the soul for GIS, which provides a range of data manipulation functions, by means of these functions, users search by or not by condition some of the entities from the original data, but also the amount of space can measure, overlay analysis, network analysis and stat. attribute data of various types entities. This study only classical spatial analysis algorithms, including:overlay analysis,polygon aggregation, line linking, cutting operations, buffer analysis.This paper search several aspects of the following:(1) study spatial data in the cutting, overlay, polygon mergers spatial analysis and other common features. (2) for the line and polygon spatial data, based on their morphological characteristics, pretreat the data valid and optimization. (3) for cutting, overlay, buffer and other spatial analysis operations in the large scale of some data and low efficiency, analyze spatial data analysis operation optimization for the induction of different forms. (4) the existence of spatial data storage management, from the index cache, data optimizationindex is proposed to optimize the program and implementation. By studying the index cache techniques to lower the cost of expanding the number of index information to improve the efficiency of data search. (5) In this paper, the distributed space computing the overall framework of parallel and distributedprocessing is a fast development of spatial data, in this paper for mass spatial data, between the calculation model, and must be able to adapt torapid development of distributed parallel computing environment.Finally, we give a parallel grid to achieve a laboratory environment, large-scale data has been better efficiency results.