Hossein Rahideh

Pressure swing adsorption is one of the most common and reasonable methods of Nitrogen gas production and has been investigated and studied by many researchers during the last few decades. Due to the cost of conducting laboratory activities and the limitation of changes in the effective parameters, the modeling of this process has been considered. With the progress and growth of simulation for estimating results such as artificial neural network, the attention of some researchers has been drawn to the application of these methods in the adsorption process and brief researches have been made in the field of predicting the behavior of Pressure Swing Adsorption with artificial neural network. The simulation of Pressure swing adsorption in four steps (pressurization, adsorption, Purge and Blow down) was carried out for the purpose of nitrogen gas separation using CMS adsorbent. The effect of CMS adsorbent performance on various variables in steady state conditions was simulated. This simulation was predicted at temperatures (293.15, 298.15 and 303.15) Kelvin and pressures (5, 6, 7) bar using the Langmuir Freundlich isotherm. Finally, the results obtaint from the Aspen adsorption software is predicted by ANN method for PSA adsorption system. The prediction results of the ANN model were agreement with the results of Aspen adsorption software. The coefficient of determination or error value (R) is close to unity. By analyzing the presented results, it was observed that the use of CMS adsorbent and Langmuir-Freundlich isotherm for nitrogen separation were well performed shown a good agreement.