Modeling of supercritical fluid extraction by hybrid Peng-Robinson equation of state and genetic algorithms

In this paper, a hybrid model using both genetic algorithms and the Peng-Robinson equation of state is developed for supercritical fluid extraction, where the genetic algorithm is used to generate the non-linear binary interaction parameter of the Peng-Robinson equation of state. Various temperatures, pressures, and solubility found in the literature are used to test the proposed model. The correlation and the mean square errors of the proposed model and the Peng-Robinson equation of state are given in the paper. The predictions of the proposed hybrid model are compared to the conventional model with a Peng-Robinson equation of state in the literature. Generally, the results using the proposed model are better than those using the conventional model because the genetic algorithm used in this paper can provide a better binary interaction parameter to fit the experimental data. The effectiveness of the proposed artificial intelligence approach is demonstrated by simulation and comparison studies.