Optimal design and modelling of sustainable bio-catalytic enzyme for wastewater treatment using response surface methodology and artificial neural network

Abstract

ABSTRACT The accessibility of various wastewater treatment systems has become increasingly prominent in current times, coinciding with a rising concern regarding the substantial content of food waste that poses a significant challenge in waste management. In this context, ecoenzymes have emerged as a promising technique for water treatment due to their remarkable efficiency in pollutant removal and cost-effectiveness. Nevertheless, many experts encounter difficulties in determining the optimal combination of ecoenzyme variables to maximize pollutant removal. The objective of this study is to apply a comprehensive approach utilizing Response Surface Methodology (RSM), Artificial Neural Networks (ANN), and Multi-objective Genetic Algorithm (MOGA) methods for the modeling and optimization of ecoenzyme-based water treatment processes. The research outcomes demonstrate a noteworthy alignment between the actual removal rates of Total Suspended Solids (TSS), Volatile Suspended Solids (VSS), and Total Ammonia Nitrogen (TAN) with the values predicted by both RSM and ANN models, validating the robustness of the regression analysis (R2 >0.95). From the optimization results, it is elucidated that to attain the maximum removal of TSS, VSS, and TAN, the treatment process should be conducted over a period of 3.5 days, employing specific enzyme concentrations of 0.07 units/ml for protease, 0.027 units/ml for amylase, and 0.36 units/ml for lipase. This study employed machine learning approaches and MOGA optimization to predict the optimum ecoenzyme properties for maximum wastewater removal utilizing the concept of a low-cost, sustainable purification technology.