Neuronal and grey modelling of milk drying in spouted bed

Abstract

The chemical composition of pastes is one of the factors responsible for the behaviour of the paste during the drying process in spouted bed. Therefore, the aim of this study was to analyse the drying process of paste in spouted bed for four types of milk: whole, skimmed, semi-skimmed and low lactose content milk, which have different concentrations of sugar and fat. A neuronal model was proposed in order to describe the influence of the presence of pastes on the spouted bed fluid dynamics. The change of moisture content of the powder during the drying process and of the heat and mass transfer phenomena were estimated using a hybrid model based on the global balance of mass and energy and a phase-coupling term given by a neural network.