modeling and simulation of water softening by nanofiltration using artificial neural network

an artificial neural network has been used to determine the volume flux and rejections of ca2+ , na+ and cl¯, as a function of transmembrane pressure and concentrations of ca2+, polyethyleneimine, and polyacrylic acid in water softening by nanofiltration process in presence of polyelectrolytes. the feed-forward multi-layer perceptron artificial neural network including an eight-neuron hidden layer has the least error in modeling this non-linear process. the overall agreement between the artificial neural network results and experimental data is very good for both the volume flux and rejections, because the maximum values of normalized bias and error are -0.01122 and 1.0737 respectively.