Research On Topology Detection And Corrosion Quantification Of Grounding Grid Based On Transient Electromagnetic Method

The grounding grid is a discharge channel for electrical equipment and protects the normal operation of the power network.The corrosion of the grounding grid seriously threatens the safety of the power system.The corrosion detection and condition evaluation of the grounding grid are particularly important.The transient electromagnetic method has been used in grounding network detection because of its non-excavation and non-stop operation characteristics,but it is mostly used for grounding network breakpoint identification.In this paper,based on the transient electromagnetic method,the topological structure detection of the grounding grid and the quantitative evaluation of the corrosion degree are carried out.The main work and conclusions of the paper are as follows:The basic principle,induction process and device mode of the transient electromagnetic method are expounded;The central loop device is selected,the calculation method of apparent resistivity is deduced,and it is used as the quantitative index of corrosion of the grounding grid.Based on the finite element method,COMSOL simulation software is used to model and simulate the topology of the grounding grid.Analyze the signals of the coil device,the current distribution of the grounding grid,and the distribution of the ground magnetic field;Introduce the magnetic gradient tensor to identify the topology of the grounding grid,which basically achieves the detection of the topology of the grounding grid;After introducing the wavelet transform edge detection method,the grid positioning accuracy of the grounding grid is improved,and the grounding grid imaging is also clearer;The experiments in the Chongqing University and the Jinjiayan 220 k V substation have verified the effectiveness of the magnetic gradient tensor and wavelet transform edge detection algorithm for the topology detection of the grounding grid.Disccuss the choice of coil position for ground grid corrosion detection,and find that the effect of quantitative detection of ground grid corrosion when the coil is above the center of the flat steel is better than the grid center;Compare the difference in magnetic field strength in the coil induced signal under different corrosion levels;Due to the nonlinear relationship between the apparent resistivity of the grounding grid and the degree of corrosion,it is proposed to use the RBF neural network with simple structure and good local mapping ability to quantify the corrosion of the grounding grid;When the quantitative analysis of the corrosion of the grounding grid is performed,the amount of data is huge.It is proposed to use PCA principal component analysis to extract the principal components of the sample data,which reduces the dimension of the sample;By inputting the extracted data into the RBF neural network,the efficiency and accuracy of the quantified calculation of corrosion have been improved.Through experiments in Chongqing University,the effectiveness of PCA-RBF neural network for quantitative detection of ground grid corrosion has been proved.