Study On Fault Line Selection Method Based On Fault Types Of Single-phase Earth Fault

The faults happened in distribution network mostly are single-phase grounding faults.When mono-criterion is used to dispose the problem of fault line detection,the accuracy rate is unsatisfied due to the complex and changeable working conditions.The method which combines mono-criterion of line selection can partly improves the accuracy,at the same time,it faces the problem of complexed data processing and high redundancy.This paper discusses the features of various single-phase earth faults,in order to classify the fault types.And then,selecting the corresponding method according to the result of fault type classification,so as to ensure the accuracy and rapidity of the fault line selection algorithm.Firstly,an arc model which is in line with the actual situation is established.Then,the characteristics of the single-phase arc grounding is analyzed.It is proposed to distinguish the arc fault from the ordinary fault by using kurtosis coefficient of the wavelet high frequency construction component of fault-phase voltage.By using the features that the wavelet high frequency energy of the transient zero-sequence current is different under different fault conditions,the ordinary faults are divided into three fault modes: strong transient process,medium transient process and weak transient process.Secondly,the simulation model of distribution network is established in ATP / EMTP.Several typical steady-state and transient methods of fault line selection are selected to analysize the working conditions of different fault angles,different transition resistances and different neutral grounding modes.And the characteristics and applicable situation of each method are studied.Thirdly,the fault measure functions are established according to the characteristics of various methods.And the fault line selection algorithm which selects the most suitable methods based on different fault types is proposed.For the conditions which must use fusion method,an improved D-S evidence theory that constructs the reliability assignment function by using Xiannong fuzzy entropy is put forward.The steps of that choosing line selection method according to fault classification are illustrated by several examples,and the superiority and fault tolerance of the method are verified by comparison.Then,the results of large amounts of data in different conditions verifes the effectiveness of the method.Finally,the protection system of the fault line selection according to fault type classification is realized by using the hardware platform of the laboratory and the virtual instrument integration environment of LabVIEW.