Computational Simulation of CO2 Sorption in Polymeric Membranes Using Genetic Programming

A statistical model based on genetic programming was developed to study the solubility of CO2 in different polymers including polystyrene, poly(vinyl acetate), polybutylene succinate and poly(butylene succinate-co-adipate). The proposed genetic model can predict the CO2 solubility with the average relative deviation of 0.095, 0.0503, 0.0312, 0.039% and R2 values of greater than 0.98. The results showed efficient applicability of the model and its outperformance in predicting the CO2 solubility compared with other modeling approaches. Therefore, the proposed model contributes to enhancing better understanding of gas/polymer systems and aids in alleviating the difficulties arising in the prediction of gas solubilities during the design and optimization of the relevant processes.