Font Size: a A A

Research On Distribution Network Reconfiguraion

Posted on:2018-05-25Degree:MasterType:Thesis
Country:ChinaCandidate:Y M ChenFull Text:PDF
GTID:2322330512482528Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
With the development of smart grid,People put forward more requirements to the distribution network in the aspects of safety,reliability and economic.Because the low voltage class of each node in the distribution network,its network losses is an important part of the power system.Distribution network reconfiguration is a distribution network automation technology,by changing the line on the state of the switch,without additional equipment.So that it can reduce network losses and improve the voltage quality.In the actual operation of the distribution network,the load is changing with time,which is difficult to deal with.In this paper,the static reconfiguration of distribution network is carried out in a single time section.Then a dynamic reconfiguration method of distribution network is proposed.Firstly,the basic theory of distribution network reconfiguration is introduced in this paper.The method of power flow calculation and the objective function of distribution network reconfiguration are given.The node hierarchical matrix and the parent node matrix are introduced to describe the hierarchical structure of the network,so that the parallel computation of the current and the voltage value is realized by combining of the forward-backward algorithm.And the minimum of system network losses is taken as the objective function to optimize the structure of the distribution network,then the corresponding constraint conditions are given.Secondly,improved quantum particle swarm optimization(IQPSO)algorithm based distribution network static reconfiguration method is proposed.Because the convergence of the quantum particle swarm optimization algorithm will be affected by the initial swarm,the Logistic mapping is introduced to improve the ergodic performance of the initial swarm.Then the adaptive adjustment of the contraction expansion coefficient is introduced to improve the dynamic adaptability of quantum particle swarm optimization algorithm.In the process of distribution network static reconfiguration,the proportion of infeasible solution is reduced by the decimal encoding method based on the basic loop.The branch loop matrix and the parent node matrix are introduced to describe the nodes hierarchical structure in the network,and the judgement method of solution in the reconstruction process is given.At last,the flow of distribution network static reconfiguration is shown.The simulation results indicate that IQPSO algorithm based distribution network static reconfiguration method is correct and effective.Finally,affinity propagation(AP)clustering algorithm based distribution network dynamic reconfiguration method is proposed.In this method,the change of distribution network load is taken into account.The researched interval is divided into several subintervals,and it considers that the load in every subinterval is unchanged.Then the data in every subinterval is clustered by AP clustering algorithm.At the same time,the isolated time section of the clustering result is smoothed,so as to achieve the purpose of reducing the switch operation number.The method divides the researched interval by the data characteristics of the node load,which can effectively avoid the influence of subjective factors.The rationality and effectiveness of the proposed method are verified by simulation.
Keywords/Search Tags:distribution network reconfiguration, power flow calculation, quantum particle swarm optimization, affinity propagation clustering algorith
PDF Full Text Request
Related items