The Geospatial Data Visualization And It’s Optimization Methods On Cluster Server

The rendering of geospatial data on the clustering server is one of the key functions of the complex geography computing platform on the new hardware architeture, it also satisfies the trend of dealing with large-scale geosaptial data in the Geographic Information System(GIS) of next generation. In the environment of clustering server, several cartographic servers which can separately complete the task of rendering constitute the cartographic clustering server connecting with each other by high-speed switch. Thus how to fully use the server resources to complete the darwing task of geospatial data together and promote the performance of server is one of the research piorities. Focusing on the geospatial data visualization and it’s optimization methods on cluster servers, this dissertation studies the following aspects.Firstly, focusing on the characteristics of the rendering of geospatial data on cluster servers, a framework of geospatial data rendering platform on the cluster server is estalished, and an online interactive mapping method basing on the hierarchical storage structure is proposed.Secondly, focusing on the problem of uneven allocation of rendering tasks between servers in the cartographic cluster server which reducing the performance of the server, a balancing alogrithm based on the rendering task assignments with dynamic load feedback mechanism of server status is proposed, which can improve the balanced results and the performance of the cartographic server.Thirdly, focusing the problem of the frequently queries with highly complexity on the same geospatial data which cost large amounts of server resources, a multi-scale management method for rendering of geographic vector data on cluster server is proposed, which can reduce the time of data preparation and the cost of IO in the process of rendering a tile image significantly. On this basis, feature simplification algorithm is applied for the data compression, which can promote the performance of rendering further while hardly affect the rendering results.Finally, a cartographic service called Hiart applied in HiGIS is designed and implemented, which indicates that all key technologies and solutions discussed in this dissertation is correct and effective.