Research On Fault Location Method Of Distribution Network Based On SR-VMD-DEO3S

Due to the low voltage level of the distribution network,the amplitude of the fault traveling wave is small,and the traveling wave will be interfered by a lot of high-frequency noise in the process of transmission and extraction,which makes it difficult to detect the fault traveling wave,and reduces the accuracy of fault location.Fast and accurate fault location technology can shorten the outage time,reduce economic losses and ensure the safe and stable operation of power system.By analyzing the characteristics of fault traveling waves,this article constructs a phase-mode transformation matrix which is not limited by the fault type,extracts the relatively stable line mode component of traveling wave,and decomposes the fault traveling wave signal into narrowband signal by variational mode decomposition(VMD).This method is more robust to noise than empirical mode decomposition and wavelet transform,and can effectively suppress modal aliasing.The teager energy operator is used to calibrate the wave head of the mode with fault features,and the three-point symmetric differential energy operator(DEO3S)is formed by smoothing the signal,which improves the noise robustness and effectively avoids the endpoint effect of Hilbert transform.Aiming at the problem that weak traveling wave signal can not be extracted effectively under strong noise background,this article proposes to introduce stochastic resonance(SR)into traveling wave fault location of distribution network,and adopted the variable scale monostable stochastic resonance model to form SR-VMD-DEO3 S fault location method of distribution network.Because the parameters of SR and VMD have great influence on the result of fault location,this article combines tabu search with particle swarm optimization,to form a tabu search-particle swarm optimization algorithm,and optimizes the parameters,which improves the convergence efficiency.Moreover,the measurement indexes to measure the effect of traveling wave de-noising and decomposition are constructed,which can accurately calibrate the time when different traveling wave arrive at the detection point,and then the fault location can be accurately obtained by using the two terminal fault location principle which is not affected by the wave speed.A 10 k V distribution network simulation model is built in MATLAB/Simulink to verify the proposed fault location method.The simulation results show that the proposed method is not limited by fault type,and is less affected by fault location and transition resistance,compared with Hilbert Huang transform and wavelet modulus maximum method,the proposed method has stronger noise robustness,higher ranging accuracy and wider application range.This thesis has 60 pictures,14 tables and 67 references.