Mathematical modeling and steady-state simulation of catalytic naphtha reforming using artificial neural networks

In this study a mathematical steady-state model for catalytic reforming of naphtha was developed. The model was applied to predict reformate composition and temperature profiles in the reforming unit consisting of three fixed bed reactors in series. The process consists of dehydrogenation, dehydrocyclization, ring expansion, paraffins and iso-paraffins cracking, naphthenes cracking, paraffins isomerization and hydrodealkylation of aromatics. Steady-state simulation of catalytic naphtha reforming process was carried out using mathematical model and then the results used for the simulation by artificial neural networks (ANNs). Finally, the results of simulation with ANNs were compared with model results, and the best performance of ANN was validated.