Maximization of Cr Removal in Continuous Counter-current Liquid-Solid Fluidized Bed: A Machine Learning Approach

Abstract

Continuous counter-current liquid solid (CCLS) fluidized bed is used as an adsorption column. Due to the counter current contacting pattern and easiness in handling the fresh as well as used adsorbent continuously, CCLS adsorption column can be preferred over packed adsorption column. In the present work experimental data for chromium removal from the wastewater using CCLS adsorption has been used for developing models using artificial neural network (ANN), a well-known machine learning technique. The percentage removal of Cr is mapped as function of liquid velocity, solid velocity, particle diameter, initial concentration, and height of the column through ANN. The developed ANN model is used as objective function for the design of the process using genetic algorithm (GA), a metaheuristic tool for optimization. The results provide specific guidelines for achieving optimum Cr removal.