Rigorous connectionist models to predict carbon dioxide solubility in various ionic liquids

Abstract

Estimating the solubility of carbon dioxide in ionic liquids, using reliable models, is ofparamount importance from both environmental and economic points of view. In this regard,the current research aims at evaluating the performance of two data-driven techniques, namelymultilayer perceptron (MLP) and gene expression programming (GEP), for predicting the solubilityof carbon dioxide (CO2) in ionic liquids (ILs) as the function of pressure, temperature, and fourthermodynamical parameters of the ionic liquid. To develop the above techniques, 744 experimentaldata points derived from the literature including 13 ILs were used (80% of the points for training and20% for validation). Two backpropagation-based methods, namely Levenberg–Marquardt (LM) andBayesian Regularization (BR), were applied to optimize the MLP algorithm. Various statistical andgraphical assessments were applied to check the credibility of the developed techniques. The resultswere then compared with those calculated using Peng–Robinson (PR) or Soave–Redlich–Kwong(SRK) equations of state (EoS). The highest coefficient of determination (R2=0.9965) and the lowestroot mean square error (RMSE=0.0116) were recorded for the MLP-LMA model on the full dataset(with a negligible difference to the MLP-BR model). The comparison of results from this model with