Research On Dynamic Reconfiguration Of Distribution Network With Distributed Generation

As a part of smart grid,distributed generation(DG)has got a larger proportion in power grid with the smart grid step by step.DG has has attracted extensive attention from researchers at home and abroad because of flexibility and no pollution.DG will affect all aspects of the power grid after it is connected to the power network,including reconfiguration which can optimize the operation of network by changing the network structure of distribution network.It can effectively improve power quality and reduce power loss of distribution network and only needs to modify the structure of the distribution network itself,without adding additional equipment.Therefore,its research is of great meaning.Firstly,the research background and current situation of distribution network reconfiguration are described in this paper.Then,the basic types of DG and its advantages and disadvantages are briefly introduced.Then,different but suitable mathematical models are established according to different types of DG.The influence of DG on distribution network is summarized.With this model,the research and analysis are carried out.This paper establishes a mathematical model with the minimum loss of network.An improved PSO is proposed.The inertia factor and learning factor of particle swarm optimization are dynamically modified.The convergence performance and optimization ability of the algorithm is improved by combining the idea of mutation in genetic algorithm.The solution space of the algorithm is compressed by decimal encoding based on loop and the network loss can be obviously declined and the voltage of each node can be increased after the reconstruction.In the aspect of dynamic reconstruction,the idea of transforming dynamic state into static state is adopted in this paper.The dynamic reconfiguration problem is decomposed into static reconstruction with constraints of the number of switch operating.The mathematical model of network loss and economic benefit is established.Time of reconfiguration is divided based on the standard deviation of node voltage of distribution network and loss of distribution network.A method of bi-level optimization to solve the problem of dynamic reconfiguration of distribution network is proposed in this paper.This method can improve the search speed of dynamic reconfiguration.Finally,the simulation is carried out by IEEE33 node system and the feasibility and validity of this method are verified.