Network Planning Of Smart Distribution System Considering Various Factors

The development of smart grid technology enables to make the most of clean, renewable energy and improve quality of power supply services. However, the access of distributed generation and new energy vehicles has an important influence on the normal operation of the distribution network and increases the difficulty and uncertainty of distribution networ k planning. This situation brings new challenges to the planning of the distribution system. In this paper, multi-objective programming model is established for distribution network with considering reliability factor. Furthermore, considering the uncertainty of the output of DG, fuzzy expected value model is set up. Lastly, the Bi-level optimum model is established for the planning of distribution network,DG and new energy vehicles, and each level deals with different planning tasks. The main work is as follows:Taking economic and reliable indicators as objectives, this disser tation presents a multi-objective planning method with considering the reliability of power supply. To improve the reliability of network, however, the radial network structure and configuration of the tie lines are optimized in the planning process. Firstly, optimal spanning tree is generated as the initial grid. Then through the iteration of multi-objective genetic algorithm optimization, initial grid is adjusted. By using the improved integer genetic code, consideration is given to the tie lines while routing the radial lines. Finally, the solutions are obtained observed as the best one of combination of economy and reliability, as well as Pareto curve. This method can obtain the optimal economic solutions under different reliability requirements, and the planner can choose flexibly in many solutions according to the actual demand. Importantly, simulation shows the effectiveness of the proposed method.In allusion to the access of DG built by users and power grid companies, this dissertation puts forward an uncertain planning method based on the fuzzy expected value. More specifically, which would be introduced so as to apply the deterministic programming methods to the fuzzy uncertai n programming of distribution network. By using the improved integer genetic code, the planning of distribution network and DG are considered at the same time. Firstly, optimal spanning tree is generated as the initial grid. Then through the iteration of g enetic algorithm optimization, initial grid is adjusted. Using the parallel optimization characteristic of the genetic algorithm, the location and capacity of distributed generation are determined in network planning process. Finally the solution with the optimal fuzzy expected value is obtained. Simulation shows the effectiveness of the proposed method.Aiming at the planning of distribution network with multiple resources such as distributed generation and new energy vehicles, a bi-level programming model is proposed. The first level is designed for the planning of the distribution network, while the planning of DG and new energy vehicles is considered at the second level. However, because of the characteristics of the different distribution network planning methods and the multiple tasks and complexity of the bi-level programming method, the first level takes the improved minimum spanning tree algorithm as method for improving the efficiency of the network planning, while the second level uses genetic algo rithm to get the optimal solution of the DG and the new energy vehicles by using the parallel optimization characteristic of the genetic algorithm. The method can get combined planning results of DG, new energy vehicle charging stations, and the distributi on network. Simulation show the effectiveness of the planning method.