The Research On Corrosion Detection And Prediction For Grounding Grids

Grounding grid plays an important role in guaranteeing the safety of the operators as well as the safe operation of the power system.But the grounding grid is always subjected to various corrosions,which make the conductor cross section reduced and the electrical performance degraded.All of these make the grounding grid become a safety hazard of the power system.To solve the problems,this paper proposes a grounding corrosion detection method,which won't affect the normal operation of the system but can quickly locate faults and prevent accidents from expanding.In addition,this paper proposes the method to predict the remaining life of the grounding grid and put forwards the formulate maintenance plans to prevent accidents.By analyzing the influencing factors and mechanisms of grounding grid corrosion and the advantages and disadvantages of existing non-destructive testing methods for grounding grids,this paper proposes a quasi-steady-state measurement method to detect the corrosion state of the grounding grid.The method uses state sensor system to simulate the corrosion of the grounding grid,analyzes the sensor equivalent circuit model,and then uses the constant potential step method to obtain the polarization resistance value.Finally,the Faraday's law is used to calculate the corrosion depth and corrosion rate of the grounding grid under different environments.In addition,because the on-site electromagnetic interference signal is more and more complex,this method uses adaptive wavelet filtering algorithm to reduce the influence of on-site interference factors.The case study proves the feasibility of the proposed method.Aiming at the problem of insufficient precision of single prediction model,this paper proposes a prediction model based on genetic algorithm optimized least squares support vector machine(LSSVM)combined with error correction.The ground grid corrosion rate sequence data has characteristics of small sample and nonlinear,while the least squares support vector machine has appliability for this type of data.LSSVM is optimized by the genetic algorithm to predict the corrosion rate sequence in this model,and the error prediction correction model can be combined to correct the prediction result,which can reduce the possibility of maximal error and further improve the accuracy and stability of the prediction model.Through comparison and verification in the case study,the model proposed in this paper has higher prediction accuracy.Based on the prediction method of residual life of analog oil and gas pipelines,this paper proposes a prediction method for the remaining life of the grounding grid.The key to achieving residual life prediction is the corrosion rate prediction and the determination of the minimum allowable thickness of the grounding grid conductor.For these two points,this paper uses the corrosion rate prediction model to achieve rate prediction,and indirectly determines the minimum thickness of the grounding grid conductor by using the grounding line thermal stability.In the case analysis,the remaining life prediction of a 220 kV substation grounding grid is realized,which provides a reference for the development of the grounding network maintenance plan.