Distribution Network Reconfiguration Based On Time Division And Improved Particle Swarm Optimization

The power system consists of the generation,transmission,transformation,distribution and consumption of the electricity.As an important part connecting transmission lines and users,the reconstruction of the distribution network changes the power flow distribution by changing the operating structure of the system,which is also an important method to improve the stability,power supply quality and economy of the distribution network.With the integration of a large number of distributed power sources,the operating status of the grid has become more and more complex,the modeling of the distribution network,the given constraint conditions,and the optimization of algorithms are facing new challenges at the same time.As a result,the research on dynamic reconfiguration of distribution network with distributed power generation is of great significance.The main contents of this paper are as follows:(1)Facing the distributed generation with new energy,this paper proposes a distribution network model which is suitable for reconfiguration and optimization.Weibull and Beta probability distribution functions are used to simulate the output curve of new energy.And,the optimization goal of minimum network loss,minimum node voltage offset,and minimum system operating cost was established.(2)Proposed a dual time division method which analyze the load curve shape based on Pearson’s correlation coefficient firstly,get the clustering of similar power curves initially.Then,Using improved fuzzy C-means cluster analysis method to divide time into several periods,convert continuous longterm dynamic reconstruction problems into static reconstruction problems at last.(3)An improved external neighborhood version of particle swarm optimization algorithm is proposed.Based on the continuity of dynamic reconstruction,every time when the distribution network is dynamically reconfigured,optimal solution of distribution network reconfiguration in the previous period is is inherited to the particle initialization and particle update in subsequent time period optimization calculations with a weight.