Research On Single-phase Grounding Fault Line Selection Method Of Distribution Network Based On VMD And ELM

At present in our country in the low voltage distribution network neutral point is widely used in the small current grounding way,its the fault types are mainly of single-phase earth fault is given priority to,although when the single-phase grounding fault occurs system can continue to maintain the normal operation of the 1 to 2 hours,but if you can't find the fault line in time,can lead to system overvoltage,reduce equipment insulation,appear even endanger personal safety accident,etc.Therefore,how to quickly and effectively find fault lines is of great significance to the stable operation of power system.Based on the analysis of the distribution network line when single-phase earth fault occurs,on the basis of transient and steady state component,and combined with the existing method of line selection,the problem of low accuracy,this paper proposes a based on the modal decomposition based on Variational modal decomposition(VMD)and Extreme learning machine(ELM)combined with the new method of fault line selection,concrete research content is as follows:Firstly,the transient zero-sequence current signals of each circuit with one period after failure were collected and decomposed successively using VMD method to obtain the Intrinsic Mode Function(IMF)components of K different modes.In this paper,a spectrum peak method is proposed to determine the value of K,and the effectiveness of the proposed method is verified by the actual fault signal simulation.Then,in order to quantitatively describe the extracted fault signal characteristics,this paper chooses to describe them from the perspectives of energy method and relative entropy algorithm respectively,and obtains two transient fault feature vectors of each circuit,so as to more effectively reflect the distinction between fault circuit and non-fault circuit.Finally ELM network are used to classify the fault characteristics of the sample processing,in allusion to the problem of the network parameters and random given here,the introduction of Artificial fish algorithm(AFSA)on the optimal choice,to get the optimal parameter combination of values,then put the training and test samples were input to the ELM,AFSA-ELM in training and testing,finally proved the effectiveness of the proposed optimization method.In order to further compare with the VMD method,the EMD method is used to decompose the fault signals under the same working conditions.Finally,the extracted fault feature samples are input into afsa-elm for training and testing,which further proves the superiority and effectiveness of the VMD decomposition method.