Comparison between the artificial neural network, SAFT and PRSV approach in obtaining the solubility of solid aromatic compounds in supercritical carbon dioxide

Prediction of degree of crystallinity for the LTA zeolite using artificial neural networks

“…The weights and biases can be adjusted by training the network using the standard back propagation algorithm. Our previous findings [29,30,[33][34][35]40] justified that designing the MLP model from normalized data is easier than working on the original data. Therefore, in the present study all the variables have been mapped between [0 1] intervals.…”

“…The neurons of each layer may interconnect to the neurons of the previous or subsequent layer(s) with respect to the type of ANN model. Various types of mathematical models can be employed as a practical tool for pattern recognition [33], function estimation [34,35], fault detection [36] and also be a powerful technique for optimization of various processes [37].…”

“…Large numbers of hidden neuron suffer from an over-fitting, and often need high computation efforts, while small numbers of hidden neuron are unable to relate dependent to independent variables with an acceptable accuracy [33][34][35].…”

Applications of SSAFT EOS for determination of the solubilities of solid compounds in supercritical CO₂.

“…Vaferi and colleagues investigated some aromatic solubility in supercritical fluid carbon dioxide by SSAFT EoS and neural network [10]. Bagheri and Shariati investigated solubility of benzoic acid in supercritical carbon dioxide using the PR and PC-SAFT EoS.…”

Mathematical-thermodynamic solubility model developed by the application of discrete Volterra functional series theory

“…The ability of assessing and analysis of models' parameters and their individual effects on the solubility are the superiority of mathematical modeling. Generally, mathematical models are classified into two groups; (i) theoretical or semiempirical equations such as models based on equations of state, and (ii) empirical equations such as density based equations [21][22][23][24][25][26].…”