Key Technology Research Of GIS Spatial Analysis Platform Based On Hadoop

The paper begins with the research background of the GIS spatial analysis and the spatialanalysis based on the cloud computing environment, in view of the current status of big GIS data,according to the principle and method of GIS spatial analysis, combine the Hadoop open sourceframework of cloud computing, with spatial join operations, buffer analysis and overlay analysisas the research object, and design the corresponding spatial analysis operation function, using theMapReduce distributed programming technology, designed several parallel spatial analysis basedon the corresponding spatial analysis operation function and built a platform for experiments andanalysis, mainly to do the following work:（1）Analysis of the current research status of GIS spatial analysis and GIS spatial analysisbased on the cloud computing environment at home and abroad, expounds the importance ofspatial analysis in GIS, the influence of the GIS data increasing for spatial analysis advantage ofcombining the the Hadoop framework of cloud computing environment and GIS spatial analysis.（2） Expounds the requirements of the Hadoop open source framework for GIS spatialdata format, designed the corresponding spatial data format conversion function based on theGIS spatial data format, implemented the conversion between space geometric data format andGeojson text data format, were used to process the input and output data format respectively.（3）In order to make the GIS spatial analysis under the cloud computing more effectively,according to the characteristics of spatial data and GIS spatial analysis, conducted the division ofspatial data and GIS spatial analysis task scheduling under the cloud computing environment,implemented the data partition based on the space filling curve and the task scheduling strategyof GIS spatial analysis, solved the load balancing problem of task scheduling.（4）Unlike general parallel GIS spatial analysis, this paper bulit up a GIS spatial analysisframework of the cloud, and design the corresponding spatial analysis operation process for thespecific spatial analysis.（5）Designed the corresponding spatial operation function for the specific spatial analysis,and designed the corresponding parallel algorithm for spatial analysis combining with thespecific spatial operation function, implemented the spatial analysis based on the cloudcomputing environment.（6）Bulit a Hadoop cloud computing platform and configured the Hive query operationdata warehouse environment,tested with different amounts of spatial data in single machine and Hadoop cluster environment,respectively,analyzed and evaluatd the experiment results of variousspatial analysis,verified the superiority of the parallel algorithm,and tested with the differentcluster scale experiment,verified the good extensibility of the clustering performance.The paper is based on the principle of the traditional GIS spatial analysis and the cloudcomputing technology,put forward the GIS spatial analysis platform design concept based on theHadoop,and designed the parallel algorithm for spatial analysis,improved the performance of theGIS spatial analysis.