【作者】 张丰；

【导师】 刘南； 刘仁义； 盛乐山；

【作者基本信息】 浙江大学 ， 地图学与地理信息系统， 2007， 博士

【摘要】 地理信息具有分布式特征，这一特征要求对空间信息的组织和处理由集中式变为分布式。传统的分布式GIS技术体系存在一些不足，特别是在海量时空信息的共享与互操作、系统的稳定性、负载均衡能力以及联机分析效率等方面，难以适应GIS大型化、社会化发展的需求。网格技术的发展为GIS的发展带来了机遇，同时也提出了挑战。本文将围绕“分布式、海量、时态属性”这一新一代GIS的主要特征，充分采用网格计算技术及与之紧密相关的Web Service、OGIS、WSRF等相关规范与支撑技术，探索适合网格计算的时空数据建模机制，并探讨GRID GIS框架下多源异构时空数据的访问、集成和互操作方法，为透明地给用户提供数据服务和功能服务，实现异构环境的功能互操作，更好地服务于海量时空数据的共享与应用提供技术支持。本论文主要的研究重点和创新点主要表现在：(1)探讨网格GIS框架体系。论述网格、空间信息网格(SIG)、Grid GIS的基本概念和基本内容，讨论了相关的网格计算、Web Service、OpenGIS、Globus、WSRF等主要支撑技术和体系。在网格GIS框架下，以面向服务的理论和技术实现为理念，探索适合网格环境的海量时空GIS服务体系和共享策略，研究应用网格技术实现时空信息服务的模式和体系结构。(2)探索面向网格的海量时空数据建模方法。分析网格环境下时空数据建模的内容，创新性地提出网格时空数据建模包括网格节点内时空数据建模和网格节点间时空数据建模。网点内时空数据模型研究重点是剖析事件连续变化发生的深层次原因，揭示要素之间时空拓扑维护和重建在时空变化描述中的作用，并基于此理论设计基于过程的动态级联事件的时空数据模型。网点间时空数据建模重在研究数据的逻辑组织，研究创新性地提出了网格缓冲池的思想，将网格节点间的时空处理转换为网格缓冲池的时空操作，以提高网格集成数据的有效利用。(3)研究网格环境下的时空数据访问、集成机制。针对网格环境下时空数据管理，引入开放网格服务结构下的数据访问与集成(OGSA-DAI)中间件，在网络服务资源框架(WSRF)体系下，扩展OGSA-DAI对空间数据操作的活动接口，实现时空数据的访问与集成。(4)搭建网格环境下的时空数据共享与互操作框架。根据广域分布式环境下网格时空数据共享机制及网格时空数据模型，利用GML对复杂几何实体、空间、时间、拓扑等的支持，采用面向对象的思想构建，继承、封装和设计GML概念模型框架，提出基于GML的时空数据表达，并以此为消息传输方式，实现基于网格服务的时空数据互操作。(5)构建网格GIS应用系统。针对浙江省土地利用资源管理的现状和存在问题，提出了基于GRID的全省土地利用资源管理新模式。在GT4环境下，建立了系统原型GridLandRS，提供面向土地利用的数据服务和功能服务，实现数据服务注册、功能服务注册以及异源时空数据的访问、集成、显示、查询、分析等功能。更多还原

【Abstract】 The character of the geographic information which is known as distributed pushes the organization and processing of the spatial data from centralization to distribution. There are weaknesses in traditional GIS technology framework, especially in the sharing and interoperation of massive spatio-temperal data and resources, stability of the system, load balance and efficiency of online analysis, which cannot meet the development of GIS anymore. The development of Grid technologies brings GIS both opportunities and challenges.According to the major characteristics of new generation GIS which is distributed, massive and tempo-attributed, spatio-temperal data modeling mechanisms for Grid computation are explored and the access, integration and interoperation of spatio-temperal data from heterogeneous environment and heterogeneous data source in Grid GIS framework is discussed based on grid compute technology and its closely related standards and supported technologies such as Web Service, OGIS, WSRF, and which is with the purpose to offer data and function service to users transparently, realize functional operation in heterogeneous environment and technologically support spatio-temperal data sharing. The focus and innovation of this dissertation are as follows:(1) Exploring the framework of Grid GIS. The basic concepts and contents of GRID, Spatial Information Grid (SIG) and Grid GIS are introduced and its related technologies of Grid computation including Web Service, OpenGIS, Globus, WSRF are discussed as well. After that, the massive spatio-temperal data service architecture and sharing strategy for grid computation are explored and the patterns and structures employing grid technology for spatio-temperal information service are studied in the grid GIS framework, which takes service-oriented theory and technology realization as target.(2) Modeling grid-oriented massive spatio-temperal data. After analyzing the content of spatio-temperal data model in grid environment, an innovational grid spatio-temperal data modeling is prompted, which include models both inside a grid node and between grid nodes. The spatio-temperal data model inside a grid node focuses on the deep reason why events occur continuously so as the role which spatio-temperal topological maintenance and rebuilding among features plays in the description of spatio-temperal changing course is exposed. As a result, a dynamic event-cascaded spatio-temperal data model process-based is designed. The spatio-temperal data model among grid nodes puts a lot focus on how to organize data logically, in which an innovational method called "grid buffer pool" is brought forward, that can convert spatio-temperal operations among grid nodes into spatio-temperal operations in the grid buffer pool in order to make full use of grid data.(3) Studying the mechanism of accessing and integrating spatio-temperal data in grid environment. In accordance with data management in grid environment, by introducing a middle-ware named Open Grid Services Architecture—Data Access and Integration (OGSA-DAI) and extending its activity to manipulate spatial data in Web Service Resource Framework (WSRF), the access and integration of spatio-temperal data is realized.(4) Establishing grid-based framework for sharing and interoperating spatio-tempral data. According to both sharing mechanism and data model of grid spatio-temperal data in distributed wide-area environment, A GML-based spatio-temperal data expression is prompted which benefited from GML’s support to complex geometry entity, space, tempo, topology and the GML concept framework which is constructed, generated, encapsulated and designed objected-oriented.(5) Developing a grid-GIS-based application. A provincial Grid-based land-use information management scheme is prompted which aims at the current state and existed problems of land-use resource management in Zhejiang province. An architecture for provincial resource management in Grid environment and spatio-temperal data model in each minimum grid node are offered, which is intended to provide both data service and function service for land-use. A system prototype that integrates the functions of service registering and management of spatio-temperal data access, integrating, displaying, statistics and overlapping analyze spatio-temperal data from heterogeneous data source is given in GT4 environment.更多还原