Artificial Neural Network (ANN) and mathematical modeling of hydration of corn cereal with milk

Abstract

The crunchiness of breakfast cereal is associated with the freshness and quality of the product, and its loss can cause rejection to consumption. So, it is important to know the hydration kinetics of different food products, as well as the influence of process conditions (such as temperature and time) on their rates. The objective of this work was to evaluate the effect of hydration on the physical-chemical properties of the milk-hydrated cereal by studying the hydration kinetics with the application of empirical models and Artificial Neural Networks (ANN). Hydration was conducted in 3 cereal/milk proportions, 3 immersion temperatures. The Peleg models was used, and the physical-chemical responses and kinetic parameters of the hydration process were considered for modeling and simulation using RNA. For the hydrated cereal, analyzes of moisture, ashes, lipids, and crude fiber were carried out. For the milk the analyzes were soluble solids and lipids. The treatments used in hydration had a significant effect (p <0.05) on all physical-chemical properties of breakfast cereal. Of the two models, the Peleg model best described the kinetics of milk absorption in the cereal. However, the use of Artificial Neural Network was more efficient in adjusting the data for absorption.