Modeling of vapor-liquid equilibrium in gas-aqueous electrolyte system using artificial neural network models

Advancing the chemical engineering fundamentals

Thermodynamics (T2-1P)

PhD Ahad Ghaemi
Iran university of science and technology
Chemical Engineering
Narmak-Tehran-Iran
Islamic Republic of Iran

Prof Mohammad Ghanadi Marageh
Atomic Energy Organization of Iran(AEOI)
Jaber Ibn Hayan Research Laboratory
Jaber Ibn Hayan Research Lab., Atomic Energy of Iran(AEOI), Tehran, Iran
Islamic Republic of Iran

Dr Shahrokh Shahhosini
Iran university of science and technology
Chemical Engineering
Narmak-Resalat-Tehran-Iran
Islamic Republic of Iran

Keywords: VLE; Electrolyte, Neural network, NH3-CO2-H2O system, Modeling

Abstract

The literature contains several thermodynamic models that are presented to model NH3 - CO2 - H2O electrolyte system. Each of these models can reasonably predict the system only in a narrow rage of temperature, pressure or concentration. In this work, multilayer perceptron (MLP) and radial basis function (RBF) artificial neural network (ANN) models were successfully used to model the system in a wide range of the conditions. The models were then validated employing an extensive vapor liquid equilibrium (VEL) database. The performance of these models to predict partial pressures of ammonia and carbon dioxide and the total pressure were evaluated by comparing their results with the experimental data, and the results of some thermodynamic models. This comparison indicates that artificial neural networks are more accurate than other reported thermodynamic models.

Presented Monday 17, 13:30 to 15:00, in session Thermodynamics (T2-1P).