Distribution Network Reconfiguration Method Based On Hybrid Optimization Algorithm

As the key technology for self-healing control of power distribution networks,network reconfiguration can optimize the network topology of the distribution network by changing the opening and closing state of the switch,which can improve the economics,safety and reliability of the power distribution system.The continuous increase in the penetration rate of distribution generation in power distribution systems has changed the distribution of power flows,network losses and power supply quality in the original distribution network.At the same time,it has also increased the uncertainty of power distribution systems and brought new challenges for the distribution network reconfiguration.This article focuses on the optimization of reconfiguration speed of distribution network,t he consideration of uncertainties of distribution generation,and the timing correlation between load and distributed power supply.There are a large number of infeasible and repetitive candidate solutions when the intelligent optimization algorithm applied to the distribution network reconfiguration.In order to solve this problem,the branch loop matrix is used to improve the encoding method.And to generate non-repetitive feasible solutions,the Euclidean distance is used to determine whether the feasibl e solution is a repetitive candidate solution.Meanwhile,the harmony search algorithm is modified,where using fish swarm preying behavior algorithm indicates directions of the candidate solutions and using chaotic catastrophe algorithm generates the cata strophic disturbance for the inferior solution to the local extremum.Therefore,the local search speed and global search capability of the algorithm are improved.The simulation results show that the proposed method was fast and effective.Aiming at the influence of the uncertainties of distribution generation output on the distribution network,an interval blind number model with credibility is proposed to weaken the influence of uncertainties.This method can refine the specific conditions of load and distributed power output intervals,and assess the credibility of each interval distribution by historical data.Using fuzzy C-means clustering,the load within a day is clustered according to its inherent similarity,then the dynamic reconfiguration of the distribution network in the time interval is converted into a static reconstruction problem in multiple periods.The simulation results of the study cases show that this method can provide multiple sets of reconstruction plans and the credibility of each set of reconstruction plans according to the influence of uncertainties,which is favorable for the deployment personnel to choose according to the actual situation.In order to fully consider the influence of the correlation and timing of load and distribution generation supply on the network reconfiguration,a distribution network reconfiguration method considering time sequence and correlation is proposed.The Cholesky factorization and the nataf transform are used to process the correlation coefficients of the input random variables.Then the sample matrix satisfying the specific distribution parameters and the specified correlation coefficients in each period can be obtained through the principle of equal probability transformation.The simulation results verify the rationality of the proposed method.