The Progressive Optimization Model on the Corrosion Rate of the Grounding Grid

Established a cross-validation of the model optimization algorithm in order to predict the corrosion rate of the grounding grid. Forward neural network input variables by principal component analysis to extract the main element, eliminating the correlation between variables; cross-validation method and change the termination of the neural network training conditions, the selection of network training model. The final simulation results show that, get a good grounding grid corrosion rate stability and generalization performance optimization prediction model can predict the corrosion rate of the grounding grid.