Research On Fault Location Of Distribution Network Based On EMTR And WT

The distribution network is an important link for the transmission of power from the transmission network to users.The safe,stable and economical distribution network can meet the needs of users.Short-circuit faults can cause huge damage to the distribution network both in terms of dynamic stability and thermal stability.If the short-circuit fault persists for a long time,it will damage the stability of the power system and even cause the power system to collapse.Therefore,it is of great research value to find the location of the fault point in time and accurately so that the maintenance department can take corresponding measures to quickly restore the power supply.The traditional fault location method has obvious limitations for solving the fault location problem of complex distribution networks.The traveling wave method needs to identify the traveling wave head.And there are too many points in the distribution network where the wave impedance is discontinuous,therefore the corresponding wave head cannot be accurately identified.The impedance method needs to solve the fault by using the corresponding model to solve the equation.The network structure and related parameters are very complicated in the distribution network,which makes it difficult to establish a suitable model.The machine learning method requires a large amount of historical data to learn and train the fixed network structure,and the network structure in the distribution network is not stable.In order to overcome the defects of traditional fault location,this paper introduces Electromagnetic Time Reversal(EMTR)theory in electromagnetics into fault location of distribution network,and uses wavelet packet denoising method to eliminate noise in fault signal in order to further improve the accuracy of fault location.Add Gaussian noise to the fault signal of the distribution network,and then use three wavelet denoising methods including wavelet packet denoising,improved threshold function denoising and multi-scale threshold denoising to test and compare.The visual effect of the waveform and the signal denoising evaluation index including signal-tonoise ratio and the square error were used to select the best wavelet denoising method.The experimental results show that the wavelet packet denoising is more suitable for processing the fault signal of the distribution network with noise.The rigorous mathematical derivation of the application of EMTR theory in fault location was extended from the double-ended EMTR fault location principle to the single-ended EMTR fault location method in the frequency domain,and then extended from the frequency domain model to the time domain.The single-ended EMTR method in the time domain was ultimately selected to analysis fault location in the distribution network,the number of observation points can be reduced and the analysis process can be simplified.In the power system simulation software PSCAD,a 9-node distribution network model was built,the waveform diagrams under various fault types were tested and the phase-mode transformation was used for decoupling analysis.The characteristics of different fault types were obtained.In the end,EMTR-WT were used to achieve high-precision fault ranging in the distribution network without being affected by fault types and transition resistance.