Research On Data Organization Model Under Multi-grid Framework

With the improvement of social economy level and people’s more deeplyperception of the objective world, the geographic information industry gets more andmore gradually attention. Under the improvement of "Digital China" and "Smart City",basic geographic information data survey and mapping project achieved significantresults. Under the new wave of information technology, the relevant governmentdepartments, such as mapping departments make a lot of effort to basic surveying andmapping work, to build a large geographic information service platform. At the sametime, every industry or trade are trying to make more informational and Intelligent."Smart Grid" and "Smart Water Network" have been gradually mature. Whichsupporting the tremendous information technology project, is the massive spatial datadescription of the objective world. As sensor technology matures, the access to volumeand accumulation of spatial data will be dramatically increased.In the big data time, big data storage, processing, transmission and sharing, etc.,have been studied in the field of computer science and many technologies have beenused such as distributed computing and storage, grid computing, NoSQL databases, WebService, and many other theories and technologies. However, in the field of geographicinformation systems, data storage is yet using the file system and relational-baseddatabase, data-processing is yet using desktop-based software, it can be seen, thegeographic information field is not ready to adapt to the big data time.Multi-grid theory is the optimization and upgrading of geographic grid. Multi-griddivides geospatial using the basic grid, you can achieve fast spatial queries and alsocorrespond it with distributed storage and computing nodes; subdivision grid torepresent geographic coordinates, unifies the spatial reference system and can be used tointegrate multi-source data. Spatial data organization and management under the framework of multi-grid, can effectively solve the multi-source data integration, sharingof spatial data using distributed storage, distributed computing of massive spatial dataand other issues. It is also an effective way and theory to deal with spatial big data in thebig data time.From the background of big data time, this paper make analysis on large volumesof spatial data, multi-source heterogeneous and complex data structures, such as thethree characteristics, and proposed requirements of spatial data organization. This papermake analysis and describes three aspects: the basic theory of multi-grid framework-geographic grid reference system theory, the core and key points of multi-grid theory,and technical basis of data organizations-NoSQL database technology. It lay thetheoretical and technology foundation of data organize structure under multi-gridframework. This paper demonstrates the selection of the spatial reference system ofmulti-grid framework and determine the method of grid segments and the division levelof the gird. Referencing of GeoHash codeing and Base32encoding method, combinedwith multi-grid division method, this paper proposed a new grid coding method. Gridcoding method in this paper reflect both multi-grid theoretical ideas, and the actualimplementation techniques to adapt. Then given the conversion algorithms ofcoordinates between grid coding and WGS-84encoding, and make analysis of theaccuracy of the conversion process issues. On the basis of the theoretical of multi-gridframework, the specific structure and the specific content of spatial data were analyzedfrom the aspects of the content and form of spatial data, and this paper determined touse NoSQL database technology as the main implementation technology of dataorganization, then use MongoDB database as the primary storage medium of spatial dataorganization and storage. Finally, spatial data sharing and distributed computing underthe framework of multi-grid are briefly analyzed, and a framework for spatial datasharing and distributed computing were built.