Font Size: a A A

The Study Of Heuristic Algorithm In Distribution Network Planning

Posted on:2004-01-03Degree:MasterType:Thesis
Country:ChinaCandidate:Z G WangFull Text:PDF
GTID:2132360095950083Subject:Power system and its automation
Abstract/Summary:PDF Full Text Request
In the planning of electric power network, several aspects have to be considered, such as the load forecasting, substation and network planning, reactive power install, reliability analysis in normal and emergency, short circuit current analysis, distribution automation, and so on.To introduce GIS platform into distribution system planning can not only enable data management convenient and enhance the work efficiency, but also make the programming work more intuitive and perspicuous. With the development of computer technology, GIS is gradually being applied into operation and management of distribution network and has a promising prospect.In this thesis, above aspects are staled firstly, distribution network planning and software development of distribution network power flow calculation and planning are then studied.This thesis composes of the following aspects:(1) Distribution network planning based on GIS and Ant Colony AlgorithmAnt Colony Algorithm is a new general-purpose heuristic algorithm for combinatorial optimization problems. The main characteristics of this method are positive feedback, distributed computation and the use of constructive greedy heuristic. The basic principle is that when ants search feeding sources, they can release secretion-pheromone on the way which they have passed. The secretion can be detected by ants and influences their behaviors. The more ants pass some routes, the more pheromone will be left on these routes. So the intensity will be enhanced continuously. Then the probability of choosing this route is higher, intensity of its is enhanced.That Ant Colony Algorithm is introduced to distribution network planning on GIS platform not only enables it convenient to consider that electric lines should be routed along the streets, but also makes he results more development. On the basis of the primary data about the distribution network development planning and the city map of construct and development of Xuchang county, the distribution network planning method based on GIS and Ant Colony Algorithm is studied.The electric lines are usually routed along streets when constructing distributionnetwork planning project. Conjunctions of the streets can be got by overriding. How to select and process the conjunctions of the streets and to connect all the loads and to assure the radiation of distribution network is a difficult problem.In this thesis, the mathematic model of distribution network planning takes the minimum sum of annual cost and penalty cost of overload as object. Spanning tree algorithm is adopted to direct search process of Ant Colony Algorithm. Optimal result can be saved in a file or displayed on the GIS platform. The feasibility and efficiency of the algorithm are verified by the result of applying it to sample systems.(2) Distribution network planning based on GIS and Tabu SearchTabu Search is a highly efficient search heuristic algorithm which has been paid close attention. It is characterized by gathering knowledge during the search and subsequently profiting from this knowledge in order to avoid trapping into local solution. In order to avoid trapping into local solution, a flexible "remember" technology has been adopted, which can record and select the planned procedure and direct the next search direction. Tabu Search contains three essential factors, which are move, Tabu list and aspiration level.In this thesis, exchange move is adopted to act on the current solution in order to generate a set of trial solutions. The principle of renewing Tabu list is FIFO. Aspiration criterion is that when Tabu move is acted on the current solution, it can get a solution which is better than all the solutions having been searched, then the Tabu move reaches the aspiration level.In this thesis, the optimal network mode takes the real-time value of capital investment into account, takes minimum annual cost as its object and feeders' capacity, the radiation of distribution network as its restrictions. The optimal result is obtained by Tabu sea...
Keywords/Search Tags:distribution network planning, geographic information system, ant colony algorithm, Tabu search algorithm, software development, object orient
PDF Full Text Request
Related items