Estimating the solubility of different solutes in supercritical CO2 covering a wide range of operating conditions by using neural network models

摘要: Abstract A new neural network-based model is proposed for solute solubility in supercritical carbon dioxide (CO2). The of fifteen solutes at different temperatures and pressures are estimated. Four network models have been tested to investigate the generalize-ability each network. results from semi-empirical models, namely Chrastil, Kumar Johnston, Bartle, Mendez-Santiago Teja, Peng–Robinson Soave–Redlich–Kwong equation state compared one estimated by using model. average absolute deviation (AAD) 5.42% those density-based representing reasonable accuracy developed estimating fluid extraction process.