Research And Practice On Key Technologies Of Electronic Chart Cloud Service

Electronic chart cloud service is an emerging technology applications designed to achieveefficient management of the global electronic chart data. It’s complex system engineering forproviding personalized, flexible and high available service. In this dissertation, a more detailedresearch and practice about the implementation of the electronic chart cloud service, data model,storage strategy, indexing mechanism, and visualization programs and other key technologieshave been done based on the cloud platform and the characteristics of the professional. The maincontents of this dissertation are as follows:1. The new features of the current development of information technology are analyzed,and the fact that geographic information software engineering no longer enjoying freeperformance boost hardware upgrade is proposed. GIS needs a new mode of thinking. Byanalyzing the background of the development of the electronic chart service, especially forchanges in the international electronic chart standard analysis, the conclusion is that a newmarine big geographic data era has been coming, and electronic chart service needs a newtechnology platform. On the basis of deep study of cloud services related concepts, electronicchart cloud service concept is proposed for the first time. And it is also the core of thisdissertation. By analyzing the research status, the GIS research features and shortcomings in thefield of high performance computing is introduced. And the contents and methods of this studycome to clear.2. By making a deep exploration of the electronic chart cloud services technical support,there are two key issues been extracted: the framework and application implementation issues.On the basis of a detailed comparison of the current cloud services platform and taking theadvantages of SC2012+L2H technology platform, the MGCS model architecture is designed.The core technology of architecture and key mechanism is analyzed. For the applicationimplementations, cloud computing environment under the service-oriented technology is studied,and the dissertation puts focus on cloud services programming model and storage structures.3. The data model of chart set theoretic in cloud environment is proposed. Based on theelements of the concept of interpretation and structure definition to make the complexoceanographic phenomenon dismantling, the common nature element of the structure is designed.The entire marine geospatial is illustrated through combining the factor structure to constitute achart set. The chart set is a collection of relational model structure and topology of the spatialstructure of the logic through analyzing chart element in the collection attribute relationship withspatial relationships. Following the framework of this theory, the combination of elements and polymerization elements in the IHO S-57standard is redefined. It is found that the chart set has187kinds of relational schema. The mode structure of the chart space set is analyzed. Theformalized representation method of point-line-surface, the set definition of geometric gathered,constraints, and operators were introduced. The implementation method of the sea set operationis introduced in cloud environment.4. The cloud storage model is designed and a reflection from global electronic chart sourcedata file set chart elements set is constructed, forming the basis of the collection of the chart datain the cloud service environment. The reflected storage of property set, space set, frameinformation set and storage of geographic features is forming s forming for chart servicesapplication. The major research is the data partition strategy of geographic features set. Theglobal grid partition based on external tangent octahedron projection and data partition algorithmbased on Hilbert set partitioning algorithm is proposed. The loading experiment is made basedon7886global electronic charts of17.8G data in4.3.1experimental environment, the load timewas17.28minutes, and the partition took9.81minutes. The effectiveness of the algorithm isvalidated by comparing the test.5. A spatial index model of geographic features in the cloud environment is proposed,including the decomposition of the index space and index rule making. The level Hilbertencoding key build Cloud-B tree index, combining with the characteristics of the cloudcomputing environment are used to establish the indexing mechanism of the main index andB-class index. The basic algorithm of space-based Cloud-B tree index query algorithms andindex node insertion and deletion algorithms is studied. Different index rule parameters affect thecollection query performance chart geographic features are tested by experiment, the result areshowed in4.5: taking the largest number of units parameter16, the most efficient index is thehead grid cells remaining at all levels of32*32, and64*64for the other levels, the averagetime in5seconds or less without scale control query.6. The global electronic chart cloud visualization is achieved, and the visualization serviceis decomposed to aggregation of data services and rendering services. The multi-scale chartelements intelligent selection algorithm is studied. It can meet the global electronic chartmulti-scale display requirements. The mapping mechanism of chart elements visualization set isproposed to support the separation of the chart data and chart symbols, the visualizationparameter and the user-specific binding. According to the type of user parameters, visualizingthe same chart set in the cloud computing environment. A scheme of the symbol parallelrendering based on visual directory with images superimposed mechanisms is designed to ensurethe performance requirements of real-time visualization. By operating80Silverlight client-sideat the same time, scheduling experiment results show that the service response time remained less than one second. The feasibility of the chart cloud visualization is validated, and a completecloud+client service platform is formed.7. The main contents and algorithms of the dissertation is proved, including the electronicchart private cloud creation, the algorithm package deployment, the electronic chart cloudservices hardware monitoring and surveillance, as well as the impact of resource nodesdynamically scaling to the set storage. A brief introduction to chart set of personalized dataservices and cloud visualization services in the B/S and C/S system is given.