Study on Prediction Model of Tch Degradation by Three -Dimensional Electrocatalysis Based on Xgboost and Mlp

Abstract

Artificial intelligence for TCH degradation prediction can reduce the cost of experiments, and research on this has attracted increasing interest. Methods and techniques on this aspect are still immature and need to be further researched. As AI grows, more and more techniques can be applied to degradation prediction. In this paper, polyaniline (PANI) was chosen as a three-dimensional particle electrode for TCH degradation. The initial concentration, voltage, electrolyte concentration, particle concentration, and time was used as input variables and the absorbance after degradation was used as output, we successfully constructed the XGB_MLP model using two methods, XGBoost and MLP, as the framework. The predictive accuracy(R2) of the model reaches 98.52% and the mean square error (RMSE) was 0.052. These two evaluation functions demonstrate the predictive ability of the proposed model. This study provides a simple, economical and efficient system for the degradation of organic pollutants, which is expected to be widely used in the field of environmental protection.

Keywords: particle electrode, PANI, XGBoost, MLP