Leveraging Artificial Intelligence Models (Gbr, Svr, and Ga) for Efficient Chromium Reduction Via Uv/Trichlorophenol/Sulfite Reaction Publisher



Sheikhmohammadi A¹ ; Khakzad P² ; Rasolevandi T³ ; Azarpira H⁴
Source: Results in Engineering Published:2025

Abstract

Due to their way better toxin evacuation proficiency, UV/trichlorophenol/sulfite progressed lessening forms have gotten more consideration in later a long time. This research mainly centers on modeling and enhancing chromium reduction in the UV/trichlorophenol/sulfite process using Gradient Boosting Regression (GBR), Support Vector Regression (SVR), and the Genetic Algorithm (GA). In this experiments, initial Trichlorophenol concentration, sulfite concentration, pH, time and initial chromium concentration were designated as X1, X2, X3, X4 and X4, respectively. It is suggested by these results that the optimization model using artificial intelligence can forecast and maintain an increase in chromium removal rate through time. The results show that the GBR model performs better than the SVR model. The training and testing data friendliness of GBR proves to have yielded a superior accuracy (measured by R² closeness to 1), and lesser errors (lower Mean Absolute Error (MAE), Root Mean Square Error (RMSE)) than SVR, implying weaker function approximations. According to the findings of this paper, it can be deduced that GBR is more accurate than SVR in approximating functions. The optimization models revealed that the most favorable variants for the given factors are X1=6.69, X2=29.29, X3=9.22, X4=87.51, and X5=2.36. Hence, according to this study AI optimization, especially when applied to the GBR model enhances the ability to eliminate this chromium in the UV/trichlorophenol/sulfite reduction process. With the two models implemented in this study, it is clear that the GBR model was more accurate than SVR and had fewer errors made than SVR. The features specified here are the best features in the reduction process AI can contribute positively to water treatment through recommending efficient ways of water treatment, cutting down on the quantity of chemicals and resources used, reducing errors and enhancing accuracy to promote sustainable ways of water treatment. © 2025 The Authors