Application of genetic programming to develop the model for estimating membrane damage in the membrane integrity test using fluorescent nanoparticle

Abstract A new approach utilizing silica fluorescent nanoparticle as a surrogate for checking the integrity of microfiltration membrane was proposed and well applied in a previous study, but the absence of a feasible estimation model for the degree of membrane damage caused that this simple membrane integrity test was not applied easily. This study proposes genetic programming (GP) as an alternative approach to develop the model to predict the area of membrane breach with other experimental conditions (concentration of fluorescent nanoparticle, the permeate water flux and transmembrane pressure). Unlike the artificial neural network that is the most common artificial intelligence technique, GP is an inductive data-driven machine learning that evolves an explicit equation with known experimental data. The results obtained with GP models evolved were satisfactory in predicting the area of the membrane breach and, with the simple membrane integrity test, the GP technique gives a practical way for estimating the degree of membrane damage. Therefore, GP could serve as a robust approach to develop an estimation model for the new membrane integrity test.