Research On Line Selection Method For Single-phase Grounding Fault In Small Current Grounding System

The probability of single-phase ground fault in distribution network is the highest,accounting for about 70% of the total faults.After the single-phase ground fault occurs in the small current grounding system,it is allowed to run with the fault for a certain period of time.During this time,the system appears three-phase unbalanced operation,and once the unstable arc grounding is formed,it will affect people's production and life and the development of the national economy.seriously,it will cause damage to national security.The small current grounding operation mode of neutral point ungrounded and neutral grounded by arc suppression coil is often adopted in distribution network in our country.Therefore,it is particularly important to accurately and effectively detect the fault lines in this system and improve the continuity of power supply.First of all,this paper analyzes the changes and characteristics of the fault characteristics of the distribution network after the single-phase ground fault occurs under the two grounding modes of the neutral point ungrounded and the neutral point grounded by arc suppression coil.The distribution network simulation model is built by using Matlab,and the simulation results are analyzed.It is clear that the fault is identified from the fault transient component of zero sequence current.Secondly,the empirical mode decomposition(EMD),which is used to separate the transient components of zero sequence current,is discussed.In order to solve the problem of end effect in EMD method,the improved mirror extension method is used to deal with it,and the simulation results show that the improvement is effective.The improved EMD method is used to decompose the zero sequence current and complete the separation of fault features.At the same time,the correlation dimension which can quantify the fault characteristics and reflect the operation state of the system is introduced,and the second derivative information method is proposed to avoid the shortcomings of subjective determination of scale-free interval.Numerical simulation is carried out on Lorenz chaotic time series.The correlation dimension,transient direction and modal energy of zero sequence current are determined,which are based on transient components.It lays a foundation for the determination of fault line selection method.Thirdly,the BP neural network with the ability of data fusion and nonlinear mapping is introduced.Aiming at the problem that BP neural network is sensitive to the initial weight and threshold and easy to fall into local optimization,genetic algorithm is used to optimize the initial weight and threshold of BP neural network to prevent BP neural network from falling into the local optimal dead zone.Finally,it is determined that the fault line selection method based on genetic algorithm to optimize BP neural network model is adopted in this paper.Finally,simulations and experiments are carried out.The fault data are collected on the traditional distribution network and the distribution network with distributed power supply,the fault characteristics are extracted,and the data samples are composed.The performance of the line selection method based on genetic algorithm to optimize BP neural network model is analyzed and compared with the traditional BP neural network.The results show that the line selection method based on genetic algorithm to optimize BP neural network model has faster convergence speed and higher accuracy.The morphological filtering and energy ratio test methods are used to eliminate the noise and extract the fault features from the field fault data of a substation in Ordos city.