Distribution Network Fault Location Method Based On Improved Empirical Mode Decomposition

Due to the separation of the power grid and the strong promotion of clean energy power generation and the country's governance of the natural environment,clean energy is now more and more common as a distributed power source.The connection of clean energy to the distribution network in the form of distributed power sources not only absorbs large-scale clean energy,but also increases the proportion of clean energy in the power industry.However,connecting clean energy as a power source to the distribution network not only leads to a complicated topology of the distribution network,but also introduces a large amount of harmonic components.When a fault occurs,the traditional line selection and positioning method will have more serious errors and some unavoidable problems,which will cause large-scale power outages and economic losses due to untimely maintenance.In order to ensure the safety of power supply and distribution,it is necessary to propose better methods and strategies to locate faults.Therefore,research and optimization of distributed power gridconnected fault location technology and power grid-connected technology are of great significance to the future development of the power grid.This paper analyzes the current status of research on fault location technology at home and abroad,and discusses the changes that will occur when the line fails due to different distributed power access locations.Therefore,the relay protection switch refuses to operate or malfunctions,resulting in excessive Large-scale power outages or failures cannot be eliminated,causing inestimable economic losses to residents' lives and industries.Therefore,this article first discusses the connection of distributed power to different locations.When a fault occurs,due to the impact of different locations on the relay protection device,and then establish a sequence network model containing distributed power,according to the three sequence network components.The calculation to explain the fault location of the distribution network will become difficult because of the addition of distributed power.The commonly used method for fault signal processing is wavelet method,but whether it is Fourier transform or wavelet transform method,it is necessary to select a fixed fundamental frequency component,on this basis,the original signal waveform is decomposed to obtain the corresponding frequency doubling component.Therefore,the waveform characteristics of the obtained frequency-doubled component are quite different from those of the original waveform,and it is inevitable that errors will occur due to the differences in characteristics during the process of positioning and line selection.The empirical mode decomposition calculation(EMD)method does not need to select the fundamental frequency component.The original new number waveform can be decomposed by calculating an optimal fundamental frequency component,and the obtained frequency-doubled component waveform characteristics and the original new number waveform Features have higher similarity.Based on this basis,MATLAB is used to build a simulation system,and the Hilbert transform empirical mode decomposition(EMD)algorithm and the improved empirical mode decomposition(MEEMD)algorithm are used to decompose the fault current of the line.The fault current component analysis finds the faulty line.From the simulation results,it can be seen that Empirical Mode Decomposition(EMD)has a high error in the accuracy of fault line selection,and has a great disadvantage in terms of calculation complexity and calculation speed.The improved modal decomposition(MEEMD)can achieve 100% accuracy from the simulation results,and it is significantly better than the empirical modal decomposition algorithm in terms of calculation speed and calculation amount.The rate is high,and the calculation speed is very fast,which is very suitable for fault line selection in distribution networks containing distributed power sources.