Dual-purpose optimization of dye-polluted wastewater decontamination using bio-coagulants from multiple processing techniques via neural intelligence algorithm and response surface methodology

Abstract

Novel Luffa cylindrica seed (LCS) extracts obtained from different processing techniques were employed for coagulation/flocculation (CF) decontamination of dye-polluted wastewater (DPW). The DPW was simulated in the laboratory using Cibacron blue dye 3GA (a reactive, azo dye) and distilled water. The bio-coagulants' proximate and instrumental characterization was performed. The duo: Response Surface Methodology (RSM) and Artificial Neural Network (ANN) models were proposed to predict color/total suspended particle (CTSP) and chemical oxygen demand (COD) removal rate using bio-coagulants. Bio-coagulant dosage, wastewater pH, and stirring time are the input variables. Based on experimental designs, RSM and ANN models have been generated. Regression coefficient (R2) and mean square error (MSE) have been implemented and correlated to test the adequacy and predictive ability of both models. The fitness of the experimental values to the expected values established that the Sutherland extract performed better. The model indicator for Sutherland extract revealed as thus: RSM (R2,0.9886 and MSE, 1.4494) for CTSP, and (R2, 0.9921 and MSE, 0.9249) for COD; and ANN (R2, 0.9999 and MSE, 0.00000057) for CTSP and (R2, 0.9999 and MSE, 0.0000000457) for COD. The obtained results revealed that ANN model was preferred for predicting the removal of CSTP and COD from DPW.