Research On Single Phase To Ground Fault Line Selection Method Of Distribution Network Based On SVD-SNN

With the improvement of national economic level,accompanied by higher demand and stability of power system.In our country,the medium and low voltage power network mostly chooses the small current grounding system.If the system is in failure state for a long time,it will lead to serious consequences,and the failure probability will directly affect the stability of the national economy.Therefore,how to accurately determine the fault location in the small current grounding system is a focus of research work.This thesis mainly focuses on the problem of rapid identification of fault line selection,and puts forward a new method of line selection,which plays a solid and stable role in the safe operation of power grid.1.The domestic and foreign scientific research achievements and the frontier development trend in the research field are analyzed in detail,and the single-phase grounding fault characteristics of the small current grounding system are listed and classified.The analysis shows that the transient zero sequence current waveform characteristics of the fault line and the non fault line are obviously different when the fault occurs.2.In order to improve the speed and accuracy of fault line selection,the twin neural network is introduced to establish the fault line selection model,and the back propagation and random gradient descent algorithm are used to optimize the parameters.In this thesis,the simulation model of small current grounding system is built,and several typical fault conditions are listed for the zero sequence current of single-phase grounding fault,The algorithm is verified by simulation.Then set different conditions of fault simulation,sort out the output fault characteristic data,as the input feature vector into the established neural network line selection model,to determine the fault line selection.3.In order to further study the applicability of svd-snn algorithm,this thesis also compares the accuracy of the other two methods.Simulation and line selection results show that the fault line selection accuracy of the proposed method is higher than that of the other two traditional similarity measure methods,namely,the Euclidean distance based method and the dynamic time bending method.