Premium
Modeling of Methane Oxidative Coupling under Periodic Operation by Neural Network
Author(s) -
Abdolahi F.,
Mortazavi Y.,
Khodadadi A.,
Hudgins R. R.,
Silveston P. L.
Publication year - 2005
Publication title -
chemical engineering and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.403
H-Index - 81
eISSN - 1521-4125
pISSN - 0930-7516
DOI - 10.1002/ceat.200407041
Subject(s) - oxidative coupling of methane , artificial neural network , methane , catalysis , biological system , coupling (piping) , selectivity , ethylene , chemistry , set (abstract data type) , control theory (sociology) , thermodynamics , computer science , engineering , organic chemistry , physics , artificial intelligence , mechanical engineering , control (management) , biology , programming language
A set of feed forward multilayer neural network models have been proposed to predict CH 4 conversion, C 2 and ethylene selectivity of methane oxidative coupling under periodic operation. These parameters predicted by the proposed neural network are based on cycle period, cycle split, and CH 4 and O 2 mole fractions in the first and second part of the period. Due to the dynamic nature of periodic operation and the kinetic complexity of the investigated reactions, the proposed approach is an effective tool to model the system. The agreement between model predictions and experimental data was quite satisfactory. The models could be employed to optimize the experimental conditions in order to get better output from the catalytic reaction. It is concluded that the neural network is an effective tool for modeling catalytic chemical reactions under periodic operation.