A Neuro-Fuzzy Model for a Dynamic Prediction of Milk Ultrafiltration Flux and Resistance

Abstract

A neuro-fuzzy modeling tool (ANFIS) has been used to dynamically model cross flow ultrafiltration of milk. It aims to predict permeate flux and total hydraulic resistance as a function of transmembrane pressure, pH, temperature, fat, molecular weight cut off, and processing time. Dynamic modeling of ultrafiltration performance of colloidal systems (such as milk) is very important for designing of a new process and better understanding of the present process. Such processes show complex non-linear behavior due to unknown interactions between compounds of a colloidal system. In this paper, ANFIS, Multilayer Perceptron (MLP) and FIS were applied to compare results. The ANFIS approximation gave some advantage over the other methods. The results reveal that there is an excellent agreement between the tested (not used in training) and modeled data, with a good degree of accuracy. Furthermore, the trained ANFIS are capable of accurately capture the non-linear dynamics of milk ultrafiltration even for a new condition that has not been used in the training process (tested data). In addition, ANFIS and Multilayer Perceptron (MLP) are compared and the Matlab software was adopted to implement the method.

Keywords

• neuro-fuzzy inference system
• milk ultrafiltration
• permeate flux
• total hydraulic resistance
• multilayer perceptron
• fuzzy inference system