Hybrid approach based on response surface methodology and artificial neural networks coupled with genetic algorithm (RSM-GA-ANN) for the Prediction and optimization for the Photodegradation of dye using nano ZnO anchored glass fiber under solar light irradiation

Abstract

BackgroundSemiconductor solar photocatalysis is an AOP methodology used for mineralizing chromatophores in waste effluent which is cost effective and sustainable nature. MethodsThe prediction of photodegradation of a model pollutant Amido black 10B dye using nanocrystalline ZnO immobilized onto glass fiber and the process was optimized using Response Surface Methodology (RSM) and Artificial Neural Networks (ANN) coupled with Genetic Algorithm (GA). The experiments for the photocatalytic degradation of AB10B dye using immobilized ZnO was designed using Box-Behnken Design (BBD) based on RSM. The variables such as pH, catalyst dosage, time, and dye concentration were designed to optimize the AB 10B dye degradation. The statistical significance of the developed model can be determined using an adjusted coefficient of regression (Adj. R2), co-efficient of regression (R2), probability value at 95% confidence interval (P-value and F-value), predicted R2, lack-of-fit, and analysis of variance (ANOVA). Significant findingsThe RSM-ANN and RSM (GA-ANN) hybrid models were developed to evaluate their performances on the prediction of the% degradation of AB 10B dye. The Regression co-efficient (R2) for the developed RSM, RSM-ANN, RSM -(GA-ANN) models were found to be 0.8672, 0.8997, and 0.9669 respectively, and revealed that the RSM-(GA-ANN) have higher prediction capability and accuracy than other two models. Hence, the developed RSM-(GA-ANN) model can be triumphantly be used for the prediction of photocatalytic process for environmental remediation with a higher degree of exactness for the industrial scale.