【摘要】：本文從GIS空間分析和云計算環(huán)境下的GIS空間分析的研究背景出發(fā)，針對當(dāng)前GIS數(shù)據(jù)量大的現(xiàn)狀，根據(jù)GIS空間分析的原理與方法，結(jié)合云計算的Hadoop開源框架，以空間連接操作、緩沖區(qū)分析和疊加分析為研究對象，設(shè)計了相應(yīng)的空間分析操作函數(shù)，運用MapReduce分布式編程技術(shù)，在相應(yīng)的空間分析操作函數(shù)的基礎(chǔ)上對這幾種空間分析進(jìn)行并行化設(shè)計，并搭建了平臺進(jìn)行實驗與分析，主要做了以下工作： （1）分析了當(dāng)前國內(nèi)外的GIS空間分析以及GIS空間分析在云計算環(huán)境下的研究現(xiàn)狀，闡述了空間分析對于GIS的重要性、當(dāng)前GIS數(shù)據(jù)量增大對于空間分析的影響以及云計算環(huán)境下的Hadoop框架與GIS空間分析相結(jié)合的優(yōu)勢。 （2）闡述了Hadoop開源框架對GIS空間數(shù)據(jù)格式的要求，根據(jù)GIS空間數(shù)據(jù)格式設(shè)計了相應(yīng)的空間數(shù)據(jù)格式轉(zhuǎn)換函數(shù)，實現(xiàn)了空間幾何數(shù)據(jù)格式與Geojson文本數(shù)據(jù)格式之間的相互轉(zhuǎn)換，分別用于處理輸入與輸出數(shù)據(jù)的格式。 （3）為了更有效的進(jìn)行云計算下的GIS空間分析，針對空間數(shù)據(jù)和GIS空間分析的特點進(jìn)行了空間數(shù)據(jù)的劃分以及云計算環(huán)境下的GIS空間分析任務(wù)調(diào)度，實現(xiàn)了基于空間填充曲線的數(shù)據(jù)劃分以及GIS空間分析任務(wù)調(diào)度的策略，解決了任務(wù)調(diào)度的負(fù)載均衡問題。 （4）與一般的并行GIS空間分析不同的是，本文搭建了云GIS空間分析框架，并針對具體的空間分析設(shè)計了相應(yīng)的空間分析操作流程。 （5）針對具體的空間分析設(shè)計了相應(yīng)的空間操作函數(shù)，并結(jié)合具體的空間操作函數(shù)設(shè)計了相應(yīng)的并行空間分析算法，實現(xiàn)了云計算環(huán)境下的空間分析。 （6）搭建了Hadoop云計算平臺和配置了Hive數(shù)據(jù)倉庫運行環(huán)境，使用了不同數(shù)據(jù)量的空間數(shù)據(jù)在單機(jī)以及Hadoop集群環(huán)境下分別進(jìn)行了實驗，對各種空間分析的實驗結(jié)果進(jìn)行了分析和評價，驗證了并行算法的優(yōu)越性，并且通過不同集群規(guī)模進(jìn)行實驗，驗證了集群性能良好的可擴(kuò)展性。 本文是在傳統(tǒng)的GIS空間分析原理與云計算技術(shù)的基礎(chǔ)上，，提出了基于Hadoop的GIS空間分析平臺設(shè)計理念，并設(shè)計了并行空間分析算法，提高GIS空間分析的性能。
[Abstract]:Based on the research background of GIS spatial analysis and GIS spatial analysis in cloud computing environment, according to the current situation of large amount of GIS data, according to the principle and method of GIS spatial analysis, combined with the open source Hadoop framework of cloud computing, this paper uses spatial connection operation. Buffer analysis and superposition analysis are used as research objects. The corresponding spatial analysis operands are designed. Based on the corresponding spatial analysis operands, the parallel design of these spatial analysis functions is carried out by using the distributed programming technology of MapReduce. And built a platform for experiments and analysis, mainly do the following work: (1) analyze the current domestic and foreign GIS spatial analysis and GIS spatial analysis in the cloud computing environment, The importance of spatial analysis to GIS, the influence of increasing GIS data amount on spatial analysis and the advantages of combining Hadoop framework with GIS spatial analysis in cloud computing environment are expounded. (2) the requirements of Hadoop open source framework for GIS spatial data format are expounded. According to the GIS spatial data format, the corresponding spatial data format conversion function is designed, and the conversion between spatial geometric data format and Geojson text data format is realized. Format used to process input and output data, respectively. (3) in order to carry out GIS spatial analysis in cloud computing more effectively, the spatial data partition and GIS spatial analysis task scheduling in cloud computing environment are carried out according to the characteristics of spatial data and GIS spatial analysis. The data partition based on the filling curve and the strategy of GIS spatial analysis task scheduling are implemented, and the load balancing problem of task scheduling is solved. (4) different from the general parallel GIS spatial analysis, this paper builds the cloud GIS spatial analysis framework, and designs the corresponding spatial analysis operation flow for the specific spatial analysis. (5) the corresponding spatial operation function is designed for the specific spatial analysis, and the corresponding parallel spatial analysis algorithm is designed in combination with the specific spatial operation function to realize the spatial analysis in the cloud computing environment. (6) build the Hadoop cloud computing platform and configure the Hive data warehouse running environment, and use the spatial data of different data amount to carry on the experiment under the single machine and the Hadoop cluster environment respectively. The experimental results of various spatial analysis are analyzed and evaluated to verify the superiority of the parallel algorithm and the good scalability of the cluster is verified by the experiments of different cluster sizes. Based on the traditional GIS spatial analysis principle and cloud computing technology, this paper puts forward the design concept of GIS spatial analysis platform based on Hadoop, and designs parallel spatial analysis algorithm to improve the performance of GIS spatial analysis.
【學(xué)位授予單位】：江西理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：P208

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 祁向前;;GIS空間分析功能在超市選址中的應(yīng)用[J];測繪科學(xué);2008年06期

2 吳亮;謝忠;陳占龍;馬麗娜;;分布式空間分析運算關(guān)鍵技術(shù)[J];地球科學(xué)(中國地質(zhì)大學(xué)學(xué)報);2010年03期

3 方金云,何建邦;網(wǎng)格GIS體系結(jié)構(gòu)及其實現(xiàn)技術(shù)[J];地球信息科學(xué);2002年04期

4 劉義;陳犖;景寧;熊偉;;基于R-樹索引的Map-Reduce空間連接聚集操作[J];國防科技大學(xué)學(xué)報;2013年01期

5 賈婷;魏祖寬;唐曙光;金在弘;;一種面向并行空間查詢的數(shù)據(jù)劃分方法[J];計算機(jī)科學(xué);2010年08期

6 陳全;鄧倩妮;;云計算及其關(guān)鍵技術(shù)[J];計算機(jī)應(yīng)用;2009年09期

7 陳優(yōu)良;翁和霞;;GIS線狀緩沖區(qū)生成算法研究[J];江西理工大學(xué)學(xué)報;2008年05期

8 韓勇,陳戈,李海濤;基于GIS的城市地下管線空間分析模型的建立與實現(xiàn)[J];中國海洋大學(xué)學(xué)報(自然科學(xué)版);2004年03期

9 羅軍舟;金嘉暉;宋愛波;東方;;云計算:體系架構(gòu)與關(guān)鍵技術(shù)[J];通信學(xué)報;2011年07期

10 劉偉;杜培軍;李永峰;;基于GIS的山西省礦產(chǎn)資源規(guī)劃環(huán)境影響評價[J];生態(tài)學(xué)報;2014年10期

相關(guān)博士學(xué)位論文 前1條

1 康俊鋒;云計算環(huán)境下高分辨率遙感影像存儲與高效管理技術(shù)研究[D];浙江大學(xué);2011年

本文編號：2397139