Study of Gas Drying by Adsorption on Composite Materials Using Neural Networks

An artificial neural network study of gas drying by adsorption in fixed bed of composite materials is presented in this paper. The experimental investigations were carried out at two values of relative humidity and three values of air flow rate respectively. The experimental data were employed in the design of the feed forward neural networks for modeling the evolution in time of some adsorption parameters, such as adsorption rate, water concentration in the bed, water vapor concentration in air at the exit from the fixed bed, drying degree and rate respectively. Based on these adsorption parameters, two composite adsorbent materials having porous matrices were compare