| With the development of electric power industry, the electric power network becomes denser, the load information varies rapidly, so do the road and constructions. There are many problems that attract people's attention, e.g. power-supply quality, environment protection, and electric de-regulation. Meantime, Geographic Information System ( GIS ) technology is developing rapidly. The GIS data management method and its application in distribution network running and planning to meet the requirements of practical distribution should be studied and established, which have important academic and practical values to improve the reliability of distribution power-supply. The distribution GIS data management methods and its application in distribution running are systematically researched, which compose distribution GIS Network Topology Analysis ( NTA ) and Optimal Rush-maintain Path ( ORP ), and distribution planning, which composes distribution Spatial Load Forecasting ( SLF ) :(1) With systematic analysis on the relationship between spatial data model and spatial data structure, the distribution feature and the two common GIS data models are analyzed, which are vector and raster data models. Then the conceptual and logic data models of distribution GIS are designed. The spatial data storage structure is given by using vector method, and their detailed data management methods are proposed. At last, the procedure of designing the property data coding and database by partitioning the dealing objects off information layer space step by step is given. Within these results, the foundation for research on its application in distribution running and planning is established.(2) The distribution electrical source and connected switch are defined to simplify the distribution GIS NTA, whose background is GIS map. Then the distribution GIS NTA is gotten, after the affected factors are analyzed, such as voltage grade. After the distribution GIS NTA coloring is defined and tested, the detailed method of realizing distribution GIS NTA coloring algorithm's function is concluded. Meanwhile, a new method that cut the rank degree of node-branch incidence matrix to distinguish radial branch from loop is proposed, which could speed up the loop topology coloring.(3) Based on the GIS network analysis function, the distribution ORP algorithm is studied. At first, the different travelers' travel time and transportation information effect on travelers' path selecting mode are analyzed, then different time segment traffic models are established by using statistical analysis method. Secondly, the practical feature of distribution faults is analyzed, then the method and its mathematical model that how to transfer the distribution ORP problem into the urban transportation shortest path problem is gotten. Thirdly, after the Dijkstra shortest path algorithm with greedy strategy is analyzed, the Dijkstra algorithm with node-arc structure is gotten, which save storage space effectively. Fourthly, from the aspect of data structure and algorithm, the improved Dijkstra shortest algorithm is proposed, which improve the traditional operations of binheap optimal array. Fifthly, from the aspect of spatial location relationship with urban transportation network feature, the practical shortest path algorithm by studying a new greedy search strategy is gotten, which is determined by the maximum degree that combined by the current node, its nearest node and destination node. At last, based on above algorithms, the distribution ORP algorithm is proposed. Test show that the proposed algorithm has the practicability and flexibility, could meet the requirements.(4) Based on the GIS spatial information analysis function, the SLF is studied. A combined SLF method to extrapolate feeder load growth is proposed by using feeder's history peak value and the merits of gray theory and Genetic Programming ( GP ). At first, load transfer coupling method to correct load history and its error for load transfer is adopted. Secondly, the real power-supply are...
|
PDF Full Text Request