多源海量地理柵格數(shù)據(jù)庫引擎技術(shù)研究
本文選題:柵格數(shù)據(jù) 切入點:多源 出處:《北京建筑大學》2013年碩士論文
【摘要】:地理柵格數(shù)據(jù)是GIS中最重要數(shù)據(jù)源之一,GIS的應用和服務越來越多地依賴于柵格數(shù)據(jù)類型。柵格數(shù)據(jù)具有數(shù)據(jù)種類繁多、數(shù)量龐大、數(shù)據(jù)格式復雜、生產(chǎn)速度快等特點,給數(shù)據(jù)管理和分發(fā)工作等帶來挑戰(zhàn)。目前很多部門使用基于文件管理地理柵格數(shù)據(jù)的方式,當數(shù)據(jù)量超過一定規(guī)模后暴露出一系列問題,嚴重影響到數(shù)據(jù)的精確檢索、高效分發(fā)和利用,降低了數(shù)據(jù)的利用率和管理效率。 針對以上問題,本論文首先從理論研究的角度,對多源海量柵格數(shù)據(jù)從其基本特征到空間數(shù)據(jù)模型再到空間數(shù)據(jù)結(jié)構(gòu)進行了詳細分析,從體系結(jié)構(gòu)、柵格數(shù)據(jù)模型和空間索引三個方面對當前主流的空間數(shù)據(jù)引擎進行了研究和對比,,重點探討了數(shù)據(jù)庫引擎的體系結(jié)構(gòu)、柵格數(shù)據(jù)存儲模型、數(shù)據(jù)庫系統(tǒng)設計、空間索引等方面的內(nèi)容;在分析柵格數(shù)據(jù)庫引擎的關(guān)鍵技術(shù)基礎上,結(jié)合開源對象-關(guān)系數(shù)據(jù)庫PostgreSQL技術(shù)設計實現(xiàn)了海量柵格數(shù)據(jù)庫引擎GeoRDE。該引擎在多個科研項目中進行了應用實踐,并取得了較好的效果。通過實踐表明,本文設計的GeoRDE具有較高的數(shù)據(jù)存儲訪問效率、高效的查詢檢索、嚴格的數(shù)據(jù)安全機制、支持多用戶并發(fā)訪問等顯著特點。 本論文的主要工作包括以下幾個方面: (1)基于柵格數(shù)據(jù)以及柵格數(shù)據(jù)庫引擎的概念,探討了柵格數(shù)據(jù)庫引擎的功能、特點及研究主要內(nèi)容,分析對比了幾種典型的空間數(shù)據(jù)庫引擎產(chǎn)品。 (2)深入研究了地理柵格數(shù)據(jù)模型和主流的空間數(shù)據(jù)模型,提出了多源海量柵格數(shù)據(jù)一體化存儲管理的總體思路和理論框架,在此基礎上設計了海量柵格數(shù)據(jù)庫引擎的總體框架,以適應海量地理柵格數(shù)據(jù)的存儲和管理;在探討柵格數(shù)據(jù)儲模型相關(guān)理論的基礎上,設計了海量柵格數(shù)的空間數(shù)據(jù)存儲結(jié)構(gòu),并建立了多源海量柵格數(shù)據(jù)庫存儲模型;詳細分析了R-Tree和GiST空間索引,結(jié)合GiST數(shù)據(jù)庫索引模板技術(shù),設計了高效的柵格數(shù)據(jù)空間索引機制,以提高海量柵格數(shù)據(jù)的查詢檢索效率;為提高柵格數(shù)據(jù)網(wǎng)絡傳輸?shù)男,設計了柵格數(shù)據(jù)壓縮策略和異步傳輸機制。 (3)對比分析了海量地理柵格數(shù)據(jù)庫引擎中的關(guān)鍵技術(shù)。分別從柵格數(shù)據(jù)組織模型及調(diào)度、空間索引機制、海量地理柵格數(shù)據(jù)的I/O優(yōu)化三個方面進行了深入的研究。提出了柵格數(shù)據(jù)分層分塊、異步傳輸、多級緩存的技術(shù)路線和實現(xiàn)思路。 (4)根據(jù)設計方案,在海量地理柵格數(shù)據(jù)庫引擎的數(shù)據(jù)組織、關(guān)鍵技術(shù)、實現(xiàn)方法等的研究基礎上,基于PostgreSQL數(shù)據(jù)庫技術(shù)設計實現(xiàn)了柵格數(shù)據(jù)庫引擎,并從數(shù)據(jù)調(diào)度、網(wǎng)絡傳輸、安全機制等方面對GeoRDE核心技術(shù)給出了完整的實現(xiàn)。最后,結(jié)合課題研制,將GeoRDE應用到多個科研項目中。研究測試結(jié)果表明,自主設研發(fā)的柵格數(shù)據(jù)庫引擎GeoRDE滿足多源海量柵格數(shù)據(jù)管理的需求。
[Abstract]:Geographic raster data is one of the most important data sources in GIS and its application and service depend more and more on raster data types.Raster data has the characteristics of various kinds of data, huge quantity, complicated data format and fast production speed, which brings challenges to data management and distribution.At present, many departments use the method of managing geographic raster data based on documents. When the data amount exceeds a certain scale, a series of problems are exposed, which seriously affect the accurate retrieval, efficient distribution and utilization of data.The efficiency of data utilization and management is reduced.In view of the above problems, this paper firstly analyzes the basic characteristics of multi-source massive raster data from its basic characteristics to spatial data model and then to spatial data structure from the perspective of theoretical research.Three aspects of raster data model and spatial index are studied and compared. The architecture of database engine, the storage model of raster data, the design of database system are discussed in detail.Based on the analysis of the key technologies of the raster database engine, a massive grid database engine, GeoRDE, is designed and implemented with the open source object-relational database (PostgreSQL) technology.The engine has been applied in many scientific research projects and achieved good results.The practice shows that the GeoRDE designed in this paper has high efficiency of data storage and access, efficient query and retrieval, strict data security mechanism, and supports multi-user concurrent access.The main work of this thesis includes the following aspects:Based on the concept of raster data and raster database engine, this paper discusses the functions, characteristics and main contents of raster database engine, and analyzes and compares several typical products of spatial database engine.(2) this paper deeply studies the geographic raster data model and the mainstream spatial data model, and puts forward the general idea and theoretical framework of integrated storage and management of multi-source and magnanimous raster data. On this basis, the general framework of the magnanimous raster database engine is designed.In order to adapt to the storage and management of massive geographic raster data, the spatial data storage structure of magnanimous grid number is designed on the basis of discussing the relevant theory of raster data storage model, and the storage model of multi-source magnanimous raster database is established.The spatial index of R-Tree and GiST is analyzed in detail, and an efficient spatial indexing mechanism of raster data is designed in combination with the technology of GiST database index template, in order to improve the query and retrieval efficiency of massive raster data, and to improve the transmission efficiency of raster data network,The raster data compression strategy and asynchronous transmission mechanism are designed.The key technologies of massive geographic grid database engine are compared and analyzed.The organization model and scheduling of raster data, spatial indexing mechanism and I / O optimization of massive geographic raster data are studied in detail.The technical route and realization idea of hierarchical block, asynchronous transmission and multilevel cache of raster data are put forward.According to the design scheme, based on the research of data organization, key technology and implementation method of massive geographic grid database engine, the grid database engine is designed and implemented based on PostgreSQL database technology, and the grid database engine is designed and implemented from data scheduling, network transmission, etc.The security mechanism and other aspects of the core technology of GeoRDE is given a complete implementation.Finally, GeoRDE is applied to many scientific research projects.The experimental results show that the Raster Database engine (GeoRDE), developed by ourselves, can meet the requirements of multi-source and magnanimous raster data management.
【學位授予單位】:北京建筑大學
【學位級別】:碩士
【學位授予年份】:2013
【分類號】:P208
【參考文獻】
相關(guān)期刊論文 前10條
1 潘志超;羅安平;張江水;;嵌入式GIS柵格數(shù)據(jù)引擎體系結(jié)構(gòu)的設計[J];測繪與空間地理信息;2010年04期
2 馬柳青;宋關(guān)福;郭會;周芹;;一種海量地形影像數(shù)據(jù)的快速漫游算法[J];地球信息科學學報;2009年05期
3 黃冬梅;張弛;杜繼鵬;高延銘;;數(shù)字海洋中海量多源異構(gòu)空間數(shù)據(jù)集成研究[J];海洋環(huán)境科學;2012年01期
4 楊春成;何列松;謝鵬;周校東;;基于空間數(shù)據(jù)庫引擎的多源地理空間數(shù)據(jù)管理[J];測繪科學技術(shù)學報;2010年05期
5 唐林燕;GeoDB的空間數(shù)據(jù)引擎[J];計算機工程;2003年13期
6 劉方敏 ,吳永輝 ,俞建新;JPEG2000圖像壓縮過程及原理概述[J];計算機輔助設計與圖形學學報;2002年10期
7 楊任農(nóng);白娟;黃震宇;鄔蒙;樊蓉;;基于SQLite的LOD模式海量影像數(shù)據(jù)管理系統(tǒng)的設計與實現(xiàn)[J];計算機工程與科學;2011年10期
8 高艷,胡啟平;數(shù)據(jù)庫通用索引樹GiST的研究[J];計算機應用;2003年06期
9 鄧錦安;王浩;;海量影像數(shù)據(jù)的組織及漫游實現(xiàn)[J];計算機與數(shù)字工程;2012年01期
10 馮玉才,胡剛;提高GiST空間利用率的方法[J];計算機工程;2003年11期
相關(guān)博士學位論文 前2條
1 蔡先華;GIS-T空間數(shù)據(jù)庫管理與應用關(guān)鍵技術(shù)研究[D];東南大學;2005年
2 謝耀華;海量影像存儲與管理關(guān)鍵技術(shù)研究[D];國防科學技術(shù)大學;2009年
相關(guān)碩士學位論文 前10條
1 閆斌;基于分布式的空間數(shù)據(jù)庫引擎設計與實踐[D];電子科技大學;2011年
2 崔鑫;海量空間數(shù)據(jù)的分布式存儲管理及并行處理技術(shù)研究[D];國防科學技術(shù)大學;2010年
3 孫琳;地理空間數(shù)據(jù)組織及調(diào)度的研究與實現(xiàn)[D];遼寧工程技術(shù)大學;2011年
4 林鍵;基礎地理空間數(shù)據(jù)管理技術(shù)方法研究[D];浙江大學;2012年
5 張雷;對OpenGIS標準的分析和空間數(shù)據(jù)引擎的實現(xiàn)[D];浙江大學;2002年
6 顧軍;R-Tree空間索引的優(yōu)化研究[D];南京師范大學;2002年
7 王晉強;基于應用模型的多源空間數(shù)據(jù)引擎研究[D];北京工業(yè)大學;2003年
8 賈鵬;柵格圖像數(shù)據(jù)庫引擎的設計與實現(xiàn)[D];中國人民解放軍信息工程大學;2005年
9 張利;多尺度海量柵格數(shù)據(jù)組織與管理的研究[D];中國人民解放軍信息工程大學;2005年
10 吳亮;基于PostgreSQL的海量數(shù)據(jù)存儲管理[D];中南大學;2005年
本文編號:1730631
本文鏈接:http://sikaile.net/kejilunwen/dizhicehuilunwen/1730631.html