Study On Multi-Objective Reconfiguration Of Active Distribution Network Based On Improved Slime Mould Algorithm

With a large number of distributed power supply connected to the grid,the electrical characteristics of distribution network have changed significantly.Active distribution network is a new development direction proposed to achieve efficient management of distribution network containing distributed power.Active distribution network reconstruction is an effective means to optimize distribution network and improve power quality.In this paper,the active distribution network with distributed power supply is taken as the main research content,and the two aspects of static and dynamic reconfiguration are deeply studied,which has certain practical significance.The specific research contents are as follows:First,this paper introduces the several kinds of mathematical model of the distributed power supply,power flow calculation method,and distribution network reconfiguration involves the knowledge of graph theory,due to the complex power grid structure make the initial population produces a large number of infeasible solution,to narrow the initial population size late can greatly accelerate the computation speed,based on graph theory and the knowledge of the minimum spanning tree algorithm,Firstly,Prim algorithm is used to generate the initial tree of continuous branch switching algorithm,and the solutions obtained through branch switching are combined with the random spanning tree solution set of Prim algorithm to eliminate the infeasible solutions.The simulation is verified by IEEE 33 and American PG&E 69 node system.Secondly,in terms of static reconfiguration,a multi-objective static reconfiguration model of active distribution network with minimum network loss,minimum voltage offset and minimum switching times is established.Pareto optimal concept is used to deal with multi-objective problems.To solve the model,slime mould algorithm was adopted.Aiming at the slow convergence speed and poor accuracy of slime mould algorithm,the search threshold and Gaussian disturbance were set,which effectively improved the convergence speed and accuracy.The refraction reverse learning strategy was introduced to jump out of the local optimal solution,and the superiority of the improved algorithm was verified by standard test function.Based on the above research,a multi-objective static reconstruction strategy of active distribution network based on improved slime mould algorithm is proposed.The simulation results show that the strategy can give the optimal solution and has certain feasibility.At the same time,the reconstruction strategy is combined with the Tai power 84 node system of the large-scale actual distribution network,which proves that it has certain practical significance.Finally,in terms of dynamic reconstruction,aiming at the uncertainty of node load and distributed power output,the dynamic reconstruction model of active distribution network was established.K-means clustering was used to divide time periods to reduce the number of reconstruction,and the dynamic reconstruction strategy of active distribution network based on improved slime mould algorithm was proposed.Simulation results show that this method can effectively reduce system network loss and improve the economy of system operation,and the validity and rationality are verified.