A Study On Optimization For Tie Lines Planning Of Urban Distribution Network Based On GIS

Urban medium-voltage distribution network, one of the most important parts of power system, is expected to distribute electric energy to consumers with considerable security and reliability. Excellent power grid structure can not only enhance capability and quality of power supply but also decrease frequency of accidents. Therefore, medium voltage distribution planning is very significant, for it directly influences future urban power grid structure as well as the reliability and economical efficiency of power network operation. Distribution network is designed in closed loop and operated in open loop. Study for tie lines planning of power distribution network is advanced in this dissertation, based on the formation of radial power network. A method for the study is proposed as follows: simple ring mode of connection should be planned primarily; along with the development of power network, more complicated modes of connection can be considered.Urban distribution network possesses obvious geographic characteristics in that transformer substations, loads and feeders are located according to geographic positions. Thus employing geographic information system can offer effective functions such as spatial search and analysis. Considering the attributes of tie lines planning, a distribution network spatial database system is developed. It is an auxiliary tool for planning, which can make planning result more valuable.Achieving the tie lines planning with minimal cost while meeting one-to-one connection between feeders is a complex global optimization problem. Based on spatial database, an algorithm for tie lines planning is presented, which is constituted by genetic algorithms and graph theory. The key idea of the algorithm is as follows: within the whole planning area, the order of traversing the transformer substations is searched by genetic algorithms; according to different orders, tie lines between the feeders of every transformer substation are planned; when planning tie lines, minimum cardinality matchings algorithm in bipartite graphs is used to achieve connection scheme between the feeders of the current transformer substation and those of its adjacent substations. Meanwhile, the cost of the tie lines must be minimized. The search range of tie lines is limited in the shortest paths among the connection points of feeders. Eventually, the optimum planning result is the one with minimal cost.According to the algorithm, software is developed to realize automatic tie lines planning. The results of examples are satisfactory. And the system can be an effective assistant tool for power distribution planning.