In the present study, genetic programming (GP) as a completely different approach in comparison with conventional methods, based on the imitation of natural evolution of living organism is proposed for the prediction of cellulose acetate (CA) polymeric membrane characteristics. The membrane preparation parameters of polymer concentration, additive concentration and coagulation bath temperature (CBT) were considered as model input parameters. GP model was developed to predict the pure water permeate flux of CA membranes based on the experimental data of a membrane pilot-scale system. Some statistical parameters were considered and calculated to investigate the reliability and accuracy of the proposed model. The model excellent general performance with R2=0.98 and the low error values confirmed the accuracy of the model. According to the results, GP has displayed the potential to be used as a reliable tool for identifying the characteristics of CA membranes.

Keywords