Spatial Index In The Power Of Gis Applications,

As information technology continues to evolve and progress, the means of processing information of people have increasingly diversified, and it contributed to the birth of intelligent enterprise management. In this background, Power GIS (Geographic Information System) is growing up.Power GIS is an intelligent analysis and management system which makes the traditional GIS technology, particularly Web GIS technology apply to the power industry and integrates the existing MIS system of the power industry. It uses hybrid approach of visual interface and text to provide optimum power distribution network model for power industry or users of various functional departments.Power GIS needs to deal with a variety of spatial data effectively, so it requires a high-performance spatial index structure. However, the spatial data of power industry has its particularity, for example, there are more point elements and line elements, but fewer polygon elements, and some individual line elements may span the whole range of geographical space. The traditional index that based on grid will make the line elements across multiple grids, so it will lead to more redundant storage and increase the storage pressure of the system. The traditional index that based on object will use the technology of Target Approximation to spatial object with approximate treatment. However, this approach will lead the line elements to distort. It is not conductive to spatial query. So it is very important to look for a new index structure which is suited for power distribution system.In order to meet the characteristics of the power distribution network, this paper has analyzed and contrasted the grid-based index with object-based index. In the end, it proposes a new mixed indexing mechanism which is suited for power industry. The indexing mechanism partitions the research region to large grid, and establishes the first index. Coarse-point objective is to reduce the number of inter-grid space objecst. Then, it establishes the second index for the space objects which are fully included in the same grid based on clusting Hilbert R-tree. Finally, it establishes indexed list for the inter-grid spatial objects. Thus, the long transmission lines of the distribution systerm span multiple grids, but its index information is stored in the indexed list. Not only to overcome the redundant storage problem in the grid-based index, but also to avoid the distortion problem of line elements in the object-based index.The second index-clustering Hilbert R tree is a new spatial index structure based on Hilbert R tree. It has made some improvements on the Hilbert R tree throngh introducting K-means clustering algorithm. The new structure makes the data sets in the same leaf node more compact and the data sets in the different leaf node overlap smaller. The clustering Hilbert R tree has better performance in the query efficiency and thus the overall performace of the mixed index has improved greatly.Finally, the indexing mechanism and the pure R tree index, and the level grid index of ArcSDE is used in the practical power GIS systerm. The compared experiment shows that the mixed indexing mechanism in practical applications more practial.