电子海图云服务关键技术研究与实践

发布时间：2018-05-22 15:37

本文选题：电子海图云服务 + 海图集合论　；参考：《解放军信息工程大学》2013年博士论文

【摘要】：电子海图云服务是一种新兴的技术应用，旨在实现全球电子海图大数据的高效管理，提供个性化、弹性化的高可用性服务，是一个较为复杂的系统工程。本文在已有云平台基础上，立足大规模电子海图网络服务的特点，对电子海图云服务的框架结构、技术方法、数据模型、存储策略、索引机制以及可视化方案等关键技术做了比较详细的研究，主要工作及成果如下： 1．提出了电子海图云服务的概念，确定了本文的研究核心。通过对当前计算机技术发展特点的分析，指出地理信息软件产品不再享受硬件提升所带来的免费性能提升，GIS的设计需要一个新的思维模式；通过对电子海图服务发展的背景分析，尤其是对国际电子海图标准的变化分析，指出一个新的海洋地理大数据时代即将来临，电子海图服务需要一个新的技术平台；在对云服务相关概念深入研究的基础上，对国内外研究现状进行了详细分析，指出了GIS在高性能计算领域的研究特点及存在的不足，明确了本文的研究目标、内容与方法。 2．创建了基于SC2012+L2H的MGCS模型架构。在详细比较当前云服务平台的基础上，指出了SC2012+L2H技术平台的优势，提炼出了电子海图云服务技术支撑的两个关键问题：框架构建问题与应用实现问题。创建了MGCS模型的云平台架构，分析了架构中的硬件虚拟化与应用虚拟化的核心技术以及服务模板管理与资源动态扩缩的关键机制；研究了云计算环境下的面向服务技术，着重分析了云服务中L2H的编程模式以及DSC存储结构的数据查询、集合创建及相关算法接口。 3．设计了云环境下的海图集合论数据模型。根据云计算环境的特征要求，通过对海图要素概念与结构的定义，实现各种复杂海洋地理现象的拆分与组合，设计了一种具有共性的要素结构，用这种要素结构的组合构成海图集合，以表达所有海图内容；通过对海图集合元素属性关系与空间关系的逻辑分析，提出了海图集合论，指出海图集合数据模型是基于笛卡尔积映射的属性模型与点集拓扑映射的空间模型的有机结合。在此理论基础上，对IHO S-57标准中的组合要素与聚合要素进行重定义，确定了海图属性集合具有187种关系模式；分析了海图空间集合的模式结构，给出了点线面及几何聚集的集合定义、约束以及操作算子的形式化表示方法，以适应云计算环境中的数据处理。 4．提出了海图集合的云存储策略。实现全球电子海图源数据文件集合到海图要素集合的映射，形成云服务环境中海图数据的基础集合；面向不同的海图服务应用，实现了属性集合、空间集合、图幅信息集合以及地理要素集合的存储映射。重点研究了地理要素集合的数据划分算法，提出了基于外切八面体投影的全球格网划分方案与基于Hilbert变阶编码的集合划分算法。在4.3.1实验环境下，对7886幅全球电子海图共17.8G数据进行了载入划分实验，载入耗时17.28分钟，，划分耗时9.81分钟，并通过对比试验，验证了算法的有效性。 5．建立了云环境下的空间索引模型，包括索引空间的分解与索引规则的制定，利用层次Hilbert编码关键字构建了Cloud-B树索引，结合云计算环境的特点建立了主索引与二级索引机制，研究了基于Cloud-B树索引的多种空间查询算法以及索引节点的插入删除等基本算法，并通过实验测试了不同索引规则参数对海图地理要素集合查询性能的影响，结果表明在4.5节的测试环境下，取最大单元数参数为16，首级网格单元数为32*32其余各级为64*64所建立的索引性能最高，不加比例尺控制的查询平均耗时在5秒以内。 6．提出了全球电子海图的云可视化服务方案，将可视化服务分解为数据服务与渲染服务的聚合，研究了多尺度海图要素的自适应选取算法，满足了全球电子海图多尺度显示的要求；提出了海图要素可视化集合的映射机制，实现了海图数据与海图符号的分离、可视化参数与具体用户的绑定，使得云计算环境下能够根据不同用户参数可视化同一海图集合；设计了基于可视化目录与图片叠加机制的符号并行渲染方案，保证了实时可视化对性能的要求；设计了客户端的并行访问实验，结果表明服务响应时间基本保持在1秒以内，验证了海图云可视化服务方案的可行性，形成了一个完整的云+端服务平台。 7．研究了云计算环境下电子海图网络服务的部署方法，包括电子海图私有云的创建、相关服务算法的模板化封装、电子海图云服务的硬件监控与服务监控，以及资源节点动态扩缩对数据集合存储影响的实验分析，并对海图集合数据服务的个性化与云可视化服务在B/S、C/S等系统中的应用做了简要介绍。
[Abstract]:The electronic sea chart cloud service is a new technology application. It aims at realizing the efficient management of the large data of the global electronic chart, providing the personalized and flexible high availability service. It is a more complex system engineering. Based on the existing cloud platform, this paper is based on the characteristics of the large-scale network service of the electric subchart, and it serves the electronic sea chart cloud service. The key technologies, such as frame structure, technical method, data model, storage strategy, index mechanism and visualization scheme, are studied in detail. The main work and results are as follows:
1. the concept of the electronic chart cloud service is put forward, and the core of this paper is determined. Through the analysis of the characteristics of the current computer technology development, it is pointed out that the geographic information software products no longer enjoy the free performance promotion brought by the hardware promotion. The design of GIS needs a new mode of thinking and the background of the development of electronic chart service. The analysis, especially the change analysis of the international electronic chart standard, points out that a new era of marine geographic data is coming, and the electronic chart service needs a new technical platform. Based on the in-depth study of the concept of cloud services, the current research status at home and abroad is analyzed in detail, and GIS is pointed out in the high performance calculation collar. The research characteristics and shortcomings of the domain have made clear the objectives, contents and methods of this study.
2. the MGCS model architecture based on SC2012+L2H is created. On the basis of the detailed comparison of the current cloud service platform, the advantages of the SC2012+L2H technology platform are pointed out, and two key problems of the electronic sea chart cloud service technology support are extracted. The framework construction and application implementation problem. A cloud platform architecture of the MGCS model is built and the framework is analyzed. The key technology of hardware virtualization and application virtualization and the key mechanism of service template management and resource dynamic expansion are discussed, and the service oriented technology in cloud computing environment is studied. The programming mode of L2H in cloud service, data query of DSC storage structure, set creation and related algorithm interface are emphatically analyzed.
3. design the data model of the set theory of sea chart under the cloud environment. According to the characteristics of the cloud computing environment, through the definition of the concept and structure of the chart elements, the resolution and combination of various complex marine geographical phenomena are realized, and a kind of common element structure is designed. By the logical analysis of the attribute relation and spatial relation of the set elements of the sea chart, a set theory of chart is proposed. It is pointed out that the set data model of the chart is an organic combination of the attribute model of the Cartesian product mapping and the spatial model of the topological mapping of the point set. On the basis of this theory, the composition elements of the IHO S-57 standard are combined with the theory. The aggregation elements are redefined, and the set of chart attribute sets has 187 relational patterns, and the pattern structure of the chart space set is analyzed. The set definition of point line surface and geometric aggregation, constraints and the formal representation of operation operators are given in order to adapt to the data processing in the cloud computing environment.
4. a cloud storage strategy for the collection of sea charts is proposed. The mapping of the global electronic chart source data files to the set of chart elements is realized to form the base set of the data of the cloud service environment, and the storage of the attributes set, space set, the set of map information and the collection of geographic elements is implemented for different sea chart service applications. This paper focuses on the data partitioning algorithm of geographic elements set, and proposes a global grid partition scheme based on external tangent eight face projection and a set partition algorithm based on Hilbert variable order coding. Under the 4.3.1 experimental environment, the total 17.8G data of 7886 global electronic charts are loaded and divided by 17.28 minutes. The validity of the algorithm is verified by a comparison experiment of 9.81 minutes.
5. the spatial index model in the cloud environment is established, including the decomposition of index space and the formulation of index rules. The Cloud-B tree index is constructed by using the hierarchical Hilbert coding key, and the main index and two level index mechanism are established in the light of the characteristics of the cloud computing environment, and a variety of spatial query algorithms and indexes based on the Cloud-B tree index are studied. In the 4.5 section of the test environment, the maximum unit number parameter is 16, the number of the first level grid unit number is 32*32 and the other levels are 64*64, and the index performance is the highest and the scale is no scale. The control query takes an average of less than 5 seconds.
6. a cloud visualization service scheme for the global electronic chart is proposed, which decomposes the visual service into the aggregation of data services and rendering services. The adaptive selection algorithm for multi-scale sea chart elements is studied, which satisfies the requirements of the multiscale display of the global electronic chart. The mapping mechanism of the visual collection of the sea diagram is proposed and the chart is realized. The separation of data from chart symbols and the binding of visual parameters to specific users makes it possible to visualize the same chart set according to different user parameters in the cloud computing environment. A symbolic parallel rendering scheme based on the superposition mechanism of visual catalogues and pictures is designed to ensure the performance requirements of real-time visualization, and the client is designed. The result of the parallel access experiment shows that the service response time is kept within 1 second, and the feasibility of the chart cloud visualization service scheme is verified, and a complete cloud + end service platform is formed.
7. research on the deployment of electronic sea chart network services under the cloud computing environment, including the creation of the private cloud of the electronic chart, the template encapsulation of the related service algorithms, the hardware monitoring and service monitoring of the electronic chart cloud services, and the experimental analysis of the dynamic expansion of the resource nodes to the data collection storage, and the collection of data clothes on the chart. Service personalization and cloud visualization services are briefly introduced in B/S, C/S and other systems.
【学位授予单位】：解放军信息工程大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：P208;U675.81

【参考文献】