PREDICTION OF THE RATE OF CROSS-FLOW MEMBRANE ULTRAFILTRATION OF COLLOIDS - A NEURAL-NETWORK APPROACH

Prediction of the dynamic crossflow ultrafiltration rate of colloids p oses a complex non-linear problem as the filtration rate has a strong dependence on both the solution physicochemical conditions and the ope rating conditions. As a result, the development of general physics-bas ed models has proved extremely challenging. In this paper an alternati ve artificial neural network approach is developed. The approach has b een used to predict the time-dependent rate of ultrafiltration of sili ca suspensions under different conditions of pH, ionic strength and ap plied pressure. Neural networks with a single hidden layer were used t o predict the filtrate flux-filtration time profiles from a small numb er of training points. Training points were chosen from both three and five sets of solution conditions to study how network predictability would be affected. Physical understanding of the process helped in cho osing the right input variables, which in turn optimised the training. The neural network approach was found to be capable of modelling this complex process accurately. (C) 1998 Elsevier Science Ltd. All rights reserved.