Prediction bubble point pressure for CO2/CH4 gas mixtures in ionic liquids using intelligent approaches

In this study, the bubble point pressure of carbon dioxide (CO2) absorption in ionic liquids (ILs) process in the presence of methane (CH4) was predicted using an artificial neural network (ANN). For this purpose, 528 data points were extracted from open literatures to develop the ANN model. These data include mole fraction of CH4 and CO2, critical properties of ILs, and the operating temperature. Levenberg-Marquardt back-propagation algorithm was used to train the optimal ANN with 6:23:1 architecture. The model findings indicated that the proposed ANN could predict precisely these kinds of systems. Regression analysis was employed to assess the performance of the developed network. The model findings indicated that the proposed ANN has the capability to predict precisely bubble point pressure of CO2 absorption in ILs solutions with average absolute relative deviation % (AARD%) of 0.25%, mean square error (MSE) of 0.0002, and correlation coefficient (R-2) of 1.0000.