Research On Fault Location Of Multi-branch Distribution Network Based On Time-frequency Information Of Traveling Wave

The accurate fault location of the distribution network has important practical significance for improving the reliability of the power supply,speeding up the processing of fault handling and recovering the power supply,and achieving efficient and economic operation of the power grid.Due to the large number of branches and complicated structure of multi-branch distribution networks,the research on fault location of multi-branch distribution lines is relatively rare.This paper proposes a method of fault location of multi-branch distribution network based on time-frequency information of traveling wave,which uses the time domain information of the main line and the frequency domain information of the branch line to carry out fault location.If the fault occurs on the trunk line,the time-domain double-ended traveling wave ranging-method is used to calculate the fault distance;if the fault occurs on the branch line,the frequency-domain single-ended traveling wave ranging method is used to calculate the fault distance.It forms a network-type multi-point measurement fault analysis method.First of all,the causes and transmission path of the fault traveling wave are analyzed.Combined with the distribution network structure,the time domain information and frequency domain information of the traveling wave under the distribution network structure are discussed according to the characteristics that the traveling wave will be folded and reflected at the discontinuous point of wave impedance.During the analysis of traveling wave time domain information,the adaptive analysis of A、C、D fault location method for multi-branch distribution network fault location is carried out.The factors that affect the inherent frequency of traveling wave are studied in traveling wave frequency domain information analysis,including the number of branch lines,the load properties at the end of branch and so on.This paper introduces the data processing methods of fault traveling wave time domain signal and frequency domain signal.In the time domain information of traveling wave,wavelet transform is used to extract the arrival time of the initial wave head,and the VMD and MUSIC spectrum method are used to extract the inherent frequency main frequency of the traveling wave when obtaining the frequency domain information.The basic principle of wavelet transform,the theory of maximum value of wavelet transform mode and the principle of VMD algorithm are introduced in detail,and the processing effect of VMD algorithm and EMD algorithm on signal is compared with that of adding white noise.According to the research of traveling wave time-frequency domain information,a multi-division dominant power network fault location based on traveling wave time-frequency composite analysis is formed in this paper.This problem is expounded from three angles:fault interval determination principle,fault interval verification and accurate fault location flow.The frequency domain characteristic matrix is formed according to the relationship between the inherent frequency of the wave in the distribution network and the transmission distance.according to the main frequency of the transient signal detected by each detection point after the fault,the real matrix of the frequency domain is formed,and the fault section discrimination matrix is constructed to determine the fault interval.If the fault occurs on the main line,the fault distance is calculated by the time domain two-terminal traveling wave distance measurement method;if the fault occurs on the branch line,the frequency domain single-terminal traveling wave distance measurement method is used to calculate the fault distance.Finally,the simulation verification is carried out,and the situation of different positions,different types of faults and different transition resistors are set up for in-depth analysis and simulation verification respectively to verify the ranging accuracy and adaptability of the fault ranging algorithm.