Design and Analysis of Artificial Neural Network (ANN) Models for Achieving Self-Sustainability in Sanitation

Abstract

The present study investigates the potential of using fecal ash as an adsorbent and demonstrates a self-sustaining, optimized approach for urea recovery from wastewater streams. Fecal ash was prepared by heating synthetic feces to 500 °C and then processing it as an adsorbent for urea adsorption from synthetic urine. Since this adsorption approach based on fecal ash is a promising alternative for wastewater treatment, it increases the process’ self- sustainability. Adsorption experiments with varying fecal ash loadings, initial urine concentrations, and adsorption temperatures were conducted, and the acquired data were applied to determine the adsorption kinetics. These three process parameters and their interactions served as the input vectors for the artificial neural network model, with the percentage urea adsorption onto fecal ash serving as the output. The Levenberg–Marquardt (TRAINLM) and Bayesian regularization (TRAINBR) techniques with mean square error (MSE) were trained and tested for predicting percentage adsorption. TRAINBR was demonstrated in our study to be an ideal match for improving urea adsorption, with an accuracy of R = 0.9982 and a convergence time of seven seconds. The ideal conditions for maximum urea adsorption were determined to be a high starting concentration of 13.5 g.L−1; a low temperature of 30 °C, and a loading of 1.0 g of adsorbent. For urea, the improved settings resulted in maximum adsorption of 92.8%.

Keywords

• optimization algorithm
• closed-loop sanitation
• human waste management
• Bayesian regularization
• Levenberg–Marquardt algorithm
• adsorption