Use of artificial intelligence in predicting the turbidity retention coefficient during ultrafiltration of water

The artificial neural network (ANN) is widely used as new mathematical tool offering an alternative way of dealing with complex problems. To predict water quality after coagulation/ultrafiltration processes using an immersed membrane, various structures of multilayer perceptrone with one hidden layer were created. Feed water turbidity, turbidity in a tank, pH and temperature in the tank as well as transmembrane pressure and permeate flux were treated as input signals. Based on the best network chosen, prognosis of water quality was done. The created ANN was able to predict very accurately behaviour of a real system with relative errors of turbidity retention coefficient equal to 2.16×10(-2).