Study Of Substation Planning Based On Intelligent Method

With the rapid development of economic and the fast course of city transition of China, the capacity of the distribution network grows very fast. Therefore the distribution network planning has been paid more and more attentions and the planning concepts of distribution network have changed greatly. In recent year, it has been gradually recognized that forecast of saturated load based on the city development planning is very important to the distribution network planning and the substation planning should be based on the forecast of saturated load. This requires that the substation location and the line corridor that is very important to the development of distribution system should be included in the city planning. Therefore, the automatic substation planning based on the spatial load forecasting has become extremely significant.This kind of substation planning is a very complex, large-scale, non-linear, multi-objective, and multi-constraint combination optimization problem. The present substation planning methods are still not efficient at performance, convergence characteristic, power supply area division and the control of transformer load factor. Based on the deep research results of former transformer substation planning, the intelligent optimization algorithm is proposed to study the substation planning problem. The main conclusions are as follows:1. Based on the research of existing substation planning method realized by voronoi diagram, a weighted voronoi diagram substation locating method (WVD) based on weight adaptive control is proposed. This method has good computation stability and overcomes the traditional method's shortcomings in substation power supply area division and load factor beyond control, etc, making the substation power supply area division more reasonable and the load factor more balanced.2. Studies in this paper show that the WVD method is very sensitive to the initial station site and the planning result is always a local optimal solution under an initial station site. According to the optimization theory′multi-point randomization global optimization strategy + issue-oriented local searching method = the most effective global optimization algorithm′, we propose the following two improved algorithms:(1) The weighted voronoi diagram substation locating and sizing method (ESGA-WVD) is realized by introducing elitist selection genetic algorithm (ESGA) to weighted voronoi diagram substation locating method (WVD).(2) The weighted voronoi diagrams substation locating method with global optimization ability based on particle swarm optimization (PSO)——PSO-WVD method is realized by integrated the global random search algorithm PSO to weighted voronoi diagrams substation locating method (WVD method).Example test results show that: ESGA-WVD method and PSO-WVD method have overcome the problem brought by only ESGA method or PSO method; PSO-WVD method, whether in running time, convergence speed or investment costs of planning result, is superior to ESGA-WVD method.3. The weighted voronoi diagrams substation locating method based on chaos particle swarm optimization——CPSO-WVD is realized by introducing chaos optimization to weighted voronoi diagrams substation locating method based on particle swarm optimization (PSO-WVD). The simulation results of uneven loads distribution examples indicate: CPSO-WVD method has overcome the precocity of PSO and well solved the global optimization of substation locating, so that the power supply area division is reasonable, meeting the requirements of each substation load factor. At the same time the examples have also confirmed the validity of CPSO-WVD algorithm.4. A divisional weighted voronoi diagram substation locating method (DWV) is proposed to investigate and simulate uneven loads distribution planning problem. The test results show that the proposed method has effectively improved weighted Voronoi diagram substation locating method (WVD).