The prediction of liquid-liquid equilibria for benzene/alkane/ionic liquids mixtures using intelligent models

Liquid-liquid extraction of aromatics from aliphatic hydrocarbons is a main process in petrochemical industry. Therefore, accurate predicting the liquid-liquid equilibria (LLE) for ternary systems of aromatic/alkane/ionic liquid can result in better liquid-liquid extraction. In this study, three intelligence methods including artificial neural network (ANN), support vector machine (SVM) and least square support vector machine (LSSVM) have been applied to predict the thermodynamic phase behavior of LLE for benzene/alkane/ionic liquid ternary systems. Optimization techniques such as particle swarm optimization (PSO), genetic algorithm (GA) and shuffled complex evolution (SCE) have been used to obtain adjustable parameters of SVM and LSSVM models. The results of prediction operation demonstrate that there was good agreement between the estimation of intelligent models and the experimental data of LLE.