Research On Intelligent Line Selection Method For Single-Phase Earth Fault In Small Current Grounding System

At present,small current grounding system is mainly used in the distribution network of domestic 3~66kV,that is,neutral point is ungrounded and arc suppression coil is grounded.In this distribution network,the incidence of single-phase-to-ground fault is very high.When the fault occurs,although the ground-to-ground voltage of the fault phase decreases,the two-phase voltage of the non-fault becomes larger and the line voltage is still symmetrical.Therefore,in a short period of time will not have an impact on the continuous supply of electricity users.However,if the single-phase-to-earth fault is maintained for a long time,the insulation will be damaged,and then the two-phase or even multi-phase short-circuit fault will occur,resulting in serious consequences.It is necessary to quickly determine the fault line and make the repair treatment.Therefore,an accurate and fast method of line selection plays a decisive role in the safe and economical operation of substation equipment and distribution network.In this paper,the current research situation at home and abroad is summarized and analyzed,and it is confirmed that fault line selection using artificial intelligence and other ideas is the future research trend in this field,which provides a train of thought for the follow-up route selection method.Then through the steady-state analysis and transient analysis of the single-phase grounding fault in the small current grounding system,the characteristics of the line before and after the fault are determined,and the criterion of line selection is found.In order to extract the fault information,this paper introduces the variational mode decomposition(VMD)method to process the fault signal,and compares it with the traditional wavelet decomposition and empirical mode decomposition(EMD)through the simulation experiment.The reliability and superiority of VMD decomposition are verified.Combined with the fault line selection criterion,a method of fault line selection by using VMD decomposition to calculate the energy difference before and after fault is proposed.In order to improve the accuracy and speed of line selection and realize the intelligence of line selection,the idea of neural network is introduced.In the aspect of algorithm,based on RBF neural network,this paper chooses wolf swarm algorithm(WPA)to optimize processing,to make up for the deficiency of RBF neural network itself.Then,aiming at the problems existing in the traditional wolf swarm algorithm,an improved wolf swarm algorithm,IWPA,is proposed to establish a VMD-IWPA-RBF neural network line selection model.Finally,this paper verifies the VMD-IWPA-RBF line selection method by establishing a simulation model.Compared with the traditional wolf group algorithm and the line selectionmethod under other algorithms,it is proved that the proposed line selection method has the advantages of higher accuracy,better speed and more precision.The paper has 30 figures,2 tables and 63 references.