Distribution Network Reconfiguration Considering Correlation Of Random Variables

Distributed generation represented by wind power generation and photovoltaic power generation and electric vehicle play an important role in allevi ating energy shortage,reducing carbon emission and harnessing haze.However,with the large-scale grid connection of DG and EV,uncertainties and related factors in the power system have gradually increased.Firstly,this paper introduces the basic principles of distribution network reconfigurat-inon,including: distribution network reconfiguration models,hierarchical forword back-ward power flow algorithm based on branch current,radial critrion of distribution network,and binary particle swarm optimiz ation algorithm applied to distribution network reconfiguration,providing the theoretical basis for subsequent chapters.Secondary,this paper analyzes and summarizes the advantages and disadvantages of various types of probabilistic flow algorithms.Considering the accuracy and speed of pr-babilistic load flow calculation,the Latin hypercube sampling Monte Carlo simulation method which can take into the correlation of random variables,is selected as the probabilistic power flow algorithm in this paper,and the program is written.this paper uses Pearson correlation coefficient to describe the correlation of random variables,adopts Nataf transform and Latin hypercube sampling method to obtain samples of random variables satisfying the specified probability distribution and correlation require-ments,and employ Monte Carlo simulation method to calculate probabilistic power flow.Based on the aboved,the influence of random variable correlation on probabilistic power flow is studied.Then,aiming at the shortcomings of the traditional balanced load model,a new distribution network reconfiguration model based on load entropy is prop osed.The simulation results in the heavy load scene of the standard IEEE33 node system show that the proposed model is more advantageous in improving the security,economy and intuit-ively judging load balance.Considering the important influence of rand om variable correlation on probabilistic power flow,a distibution network reconfiguration model based on chance constrained programming is further proposed,studying impacts of random variable correlation on the distribution network reconfiguration with the goal of reducing network loss,balancing load,and improving voltage quality.The simulation results in the improved IEEE33 node system show that the random variable correlation affect s the distribution network reconstruction result with the goal of r educing network loss and balancing load,and there is no impact on the reconstruction result with the goal of improving voltage quality.Finally,due to the real-time changes of load,DG,and EV output in the distribution system,the static reconstruction based on a certain time section cannot meet the real-time requirements.This paper adopts a dynamic physical optimization strategy to dynamically reconfigure the distribution network.Firstly,the time interval is divided into a number of time periods,and a static reconstruction considering the correlation of the random variables is performed at each time interval,and then preliminary integration of the time period is performed according to the reconstruction results.Secondary the benefit evaluation index is defined to guide secondary consolidation of time periods.Finally,this paper deals with the switch constraint problem and completes the final merge of time periods to get a dynamic reconstruction scheme of the distribution network.The simulation results show that this strategy can obtain an ideal feasible solution for the dynamic reconstruction of distribution network under the condition of satisfying the switch constraints.The conclusion of this paper can promote the distribution network reconfigu ration decision more in line with the actual project.