Research On Fault Feature Extraction And Diagnosis Method Of Grounding Grid

Fault diagnosis of grounding grid is a significant measure to ensure the normal operation of power system.In engineering practice,the fault diagnosis of grounding grid still stays in the original stage of sampling excavation.The method has characteristic of blindness,heavy workload and slow speed,is limited by the field operating conditions,so it can not accurately determine the degree of corrosion.Grounding grid fault has become a major hidden danger of the safe operation of the power grid,which will lead to the deterioration of the grounding grid performance,and will be a serious threat to the power equipment and personal safety.To find a simple and accurate method for the fault diagnosis of grounding grids in the case of power failure and no excavation,it has become an urgent task for the power sector.In this paper,the article firstly summarized the research status of fault diagnosis theory of grounding grid,and the fault feature extraction method based on empirical wavelet transform is proposed.By extracting the fault signal frequency maxima adaptively split Fourier spectrum isolated from different modes,the Fourier spectrum of fault signals are divided,and filtering the spectrum division.Then we can get the instantaneous amplitude and frequency sense,accurately characterize the grounding fault of high frequency signal,effectively reveal the frequency structure the fault data,and directly indicate the severity of grounding grid corrosion.In the meantime,a high frequency model of grounding grid is built to verify the effectiveness of the proposed method.It can be used to represent the health status of grounding grid by quantizing the frequency response signal.On this basis,the article studyed the application of quantum particle swarm optimization algorithm in fault diagnosis of grounding grid is studied.Based on the frequency-domain fault response calculation of the Pearson Correlation Coefficient of signal acquisition,the health of different grounding fault threshold,the identification and classification of the relevant vector machine,can be a very good combination of single data,complete the quantitative classification of the grounding fault.The method is suitable for real-time monitoring of grounding grid.In order to realize the accurate diagnosis of grounding location,the airticle initially studied the possibility of the high density resistivity method in grounding breakpoint diagnosis,and constructs the model of grounding grid containing soil parameters,using the least squares inversion of soil resistivity,the simulation results verify the feasibility of this method,which provides a new theoretical reference for fault diagnosis of grounding grid.