Study On Single-phase Ground Fault Of Non-direct Grounding System

In a distribution network where the neutral point is not grounded,the electromagnetic voltage transformer installed at the busbar will often “induced” the ferromagnetic resonance phenomenon due to the disturbance of the system.The power frequency ferromagnetic resonance will cause the system zero-sequence voltage rise.High,the same as the single-phase ground fault that occurs frequently in the distribution network,causing the system to have an illusory grounding phenomenon,which leads to the staff mistakenly thinking that there is a single-phase ground fault,causing the system to misjudge,so it is necessary to accurately identify the system disturbance and Ferromagnetic resonance and single-phase ground fault caused by other faults.Firstly,the steady-state characteristics of single-phase grounding and the fault characteristics of ferromagnetic resonance are analyzed,and the causes of PT "induced" ferromagnetic resonance phenomenon are expounded.By comparing and comparing the similarities and differences between the zero-sequence voltage and the zero-sequence current of the two faults,the phase difference between the zero-sequence voltage and the zero-sequence current of each feeder after the two faults are selected is analyzed as the basis for fault identification.In the calculation of the phase difference,the Hilbert transform method is introduced to solve the phase of the zero-sequence voltage and the zero-sequence current,and the fault identification result is obtained by the phase difference result obtained after the transformation.Then,for the problem of single-phase ground fault line selection,the paper summarizes the characteristics and deficiencies of the domestic and international line selection methods;according to the occurrence of single-phase ground fault,the steady state process has a long duration but a small amplitude,the duration of the transient process The short but large amplitude characteristic determines the method of combining the steady state information and the transient information in the fault line selection,using Empirical Mode Decomposition(EMD)and Fast Fourier Transform(Fast Fourier Transform).,FFT)extracts the transient energy,active power component,and fifth harmonic characteristic components in the zero-sequence current signal,and combines these three characteristic components as fault features to overcome the shortcomings of the single line selection method.Finally,the basic theory of Least Squares Support Vector Machine(LSSVM)and classification learning algorithm are described to determine that it can be applied to the fault line selection of small current grounding system.Since the parameters of LSSVM directly determine the accuracy of line selection,in order to improve the performance of LSSVM,the paper introduces the characteristics of Particle Swarm Optimization(PSO),which is simple,effective and easy to implement,and optimizes the parameters of LSSVM.Optimized Least Squares Support Vector Machine(PSO-LSSVM)line selection method.