Research On Corrosion Protection And Prediction Method Of Grounding Grid

The grounding grid is one of the equipment that must be installed in a variety of power facilities,and is mainly used to ensure the safety of personnel and other equipment.With the rapid development of power systems towards higher voltages and larger capacities,higher requirements are put forward for the safety and reliability of system operation,and the performance requirements for grounding grids are also getting higher and higher.Affected by national conditions,ordinary carbon steel is generally used for grounding grids in my country.Such materials are prone to corrosion in the soil environment,which reduces the performance of the grounding grids and even fails to work properly,leading to serious accidents.Therefore,how to avoid such problems has become the focus of research by relevant personnel.This dissertation studies the countermeasures of grounding grid corrosion from two directions.On the one hand,in order to slow down the corrosion process,the grounding grid is protected by studying the preparation of corrosion-resistant coatings.Taking advantage of the good corrosion resistance and mechanical properties of epoxy resin and asphalt materials,by adding micro-scale fillers(graphite powder and expanded graphite)and nano-scale fillers(multi-walled carbon nanotubes)to obtain a layered structure,optimize the conductive filler Adding ratio to prepare a coating material with good conductivity.By giving the coating material super-hydrophobicity,the corrosion effect of the electrolyte solution in the soil on the grounding grid is reduced.Since the superhydrophobic surface usually does not have good mechanical strength,the filler is evenly dispersed in the resin/asphalt to obtain a self-similar structure,so that after the coating is worn out,the new surface still has good superhydrophobic properties.Through microstructure analysis,super hydrophobic experiment,conductivity experiment,wear experiment,corrosion resistance experiment,self-cleaning experiment and thermal stability experiment,further test the performance of the coating material.On the other hand,by establishing a corrosion rate prediction model,the operation status of the grounding grid can be grasped in time to avoid accidents.Aiming at the shortcomings of the BP neural network model,in order to improve efficiency and avoid falling into the local optimal solution,the cuckoo algorithm is used to optimize the initial weights and thresholds of neurons,and the BP neural network model optimized by the cuckoo algorithm is established to realize the grounding network.Prediction of corrosion rate.Since the prediction accuracy of the support vector machine model is greatly affected by the initial parameter values,the penalty factor and kernel function parameters are optimized by combining the cuckoo algorithm,and a grounding grid corrosion rate prediction model based on the cuckoo algorithm optimized support vector machine is established.Taking into account the small sample and nonlinear characteristics of the grounding grid corrosion problem,in order to reduce the impact of overfitting,an extreme learning machine model with the ability to quickly deal with nonlinear relationships is used,and the particle swarm algorithm is used to optimize its initial input weight and threshold.The prediction model of grounding grid corrosion based on particle swarm optimization extreme learning machine.The prediction performance of the three models was tested separately through the existing grounding grid corrosion data,and compared and analyzed with other prediction algorithms.