Application of adaptive neuro-fuzzy inference system as a reliable approach for prediction of oily wastewater microfiltration permeate volume

In this paper, ANFIS (adaptive neuro-fuzzy inference system) as a powerful tool for modeling complex and nonlinear systems, was used to predict permeate volume of oil/water membrane separation process. The data used for modeling the flux behavior consisted of three inputs (TMP, oil concentration, filtration time) and experimental permeation values as the output. First type gaussian membership function was used for fuzzification of input variables and hybrid algorithm was chosen for the learning method of input–output data. Very well agreements were observed between experimental and simulation results. From the results, the ANFIS can be used as a reliable tool for prediction of microfiltration systems’ behavior. The coefficient of determination (R2) between the experimental and predicted values was greater than 0.99 and the mean percentage error was less than 2%, showing the great efficiency and reliability of the developed model.