Study On The Fault Line Detection For Small Current Grounding System

Currently,the medium and low voltage distribution network is mostly small current grounding system in our country.In the small current grounding system,single-phase to ground fault accounts for the majority.In recent years,many methods have sprung up,but the effect is not ideal.Fault line selection is still a difficult problem in distribution network,and we need to study more deeply.A new method of fault line selection for small current grounding system based on variational mode decomposition is proposed in this paper.In this paper,the transient state characteristics and the zero sequence network of single-phase grounding fault are analyzed.Variational mode decomposition energy entropy and optimized K-means clustering are used to research on fault line selection.Firstly,the article introduces the theory of VMD and the simulation signal is decomposed into sub-modal with the main frequency component by using VMD.From the decomposition results,the difference between the decomposed sub modes and the original signal is very small.It is concluded that VMD can decompose the signal more accurately in time range.By comparing the modal spectrum and the center frequency after decomposition,it can be concluded from the simulation results that the decomposition effect of VMD is very good in the frequency domain.The simulation results show that the VMD decomposition results in the virtual component caused by modal aliasing and over decomposition due to the different number of submodes decomposed.Secondly,in order to further extract fault information,energy entropy theory is introduced.Entropy is a function used to describe the complexity of a system.The entropy of a system represents its uniformity.The VMD method and energy entropy are considered together,and the entropy of sub modality is calculated as the criterion of line selection.Finally,this paper introduces the optimized K-means clustering algorithm,and applies the variable step FA algorithm to optimize the traditional K-means.The variable step size coefficient is introduced into the FA algorithm,which is of great significance for improving the efficiency and accuracy of FA.When the variable step size starts to be optimized,the firefly’s moving span is large,’so that it can search for larger areas and improve the speed of search.With the change of time,the later optimization process tends to be stable,and the span of firefly moving is smaller.At this time,variable step size can be used for more precise optimization.Firefly Algorithm of the variable step is applied to optimize the traditional K-means,which solves the local optimum problem caused by the selection of initial cluster centers.It is divided into two cluster centers to select the faulty line and the normal line,so as to realize the choice of the fault line.A simulation model of small current grounding system based on Simulink is built.The model consists of 6 cable mixed lines,using the actual data of substations in a certain area of Shandong province.Simulation experiments are carried out in order to verify the validity of the fault line selection method proposed in this paper,such as different line faults,different fault locations,different grounding resistances.