Research On Algorithm Of Single-phase-to-ground Fault Line Selection For Small Current Neutral Grounging System

Most of the faults in small current neutral grounding system are single-phase grounding faults.Thus,long-term faulty operation poses a great threat to power system safe operation.However,owing that the structure of China's power system is complex and inconvenient,and the characteristic signal is weak during the fault time.most of the fault line selection methods have problems.such as,slow line selection speed and low line selection accuracy.If only one line selection algorithm is used to determine the fault line,the error of line selection result may be large due to the limitation of the method itself.Therefore,studying the single-phase ground fault line selection method in small-current grounding systems has certain application value and research significance.A fault line selection algorithm based on information fusion is used.Firstly,the steady-state characteristics and transient characteristics of the neutral point are analyzed when single-phase ground fault occurs in two operating modes.The fault current,fault voltage and other fault parameters are calculated in detail,which are employed as the theoretical basis for fault line selection.Secondly,through the comparison of various fault line selection methods,three traditional line selection algorithms,namely transient energy method,fifth harmonic method and wavelet packet analysis method,are picked out.The line selection criteria for each algorithm are constructed accordingly,in which MATLAB is used to verify the effectiveness in different operating modes for neutral.Simulation results show that the single line selection algorithm is greatly affected by the fault grounding conditions,resulting in lower accuracy of fault line selection results.Then,through the study of information fusion,the working principle and training process of LVQ neural network.an information fusion method based on LVQ neural network is determined.Finally,Through the simulation model is built in MATLAB/Simulink platform,the experimental data was obtained in different operating modes for neutral.The neural network model was trained and tested using the obtained data to verify the validity of selection method.It is proved that the intelligent algorithm that combines features of multiple line selection methods can significantly improve the accuracy of fault line selection.Simulation results validate the effectiveness of the proposed method.It improves the efficiency of the maintenance personnel in finding the faulty line,and helping restore the power supply as soon as possible.Hence.the proposed method has certain practical value.