Prediction Of Permeate Flux Decline In Crossflow Membrane Filtration Of Colloidal Suspension: A Radial Basis Function Neural Network Approach

Abstract

The capability of a Radial Basis Function Neural Network (RBFNN) to predict long-term permeate flux decline in crossflow membrane filtration was investigated. Operating conditions such as transmembrane pressure and filtration time along with feed water parameters of particle radius, solution pH, and ionic strength were used as inputs to predict the permeate flux. Simulation results indicated that one single RBFNN accurately predicted the permeate flux decline under various experimental conditions of colloidal membrane filtrations and produced better predictability compared to those of the multiple regression method and regular multi-layer feed-forward Back Propagation Neural Network (BPNN), due to the more sophisticated training system of RBFNN. Further development of the artificial neural network approach to membrane filtration will enable the design of full or large scale processes with lab or pilot scale experiments.