Modeling of heat and entropy sorption of maize (cv. Sc704): neural network method

Abstract

Equilibrium moisture content of maize affects its values of dehydration heat and entropy. Precise prediction of heat and entropy with regard to its equilibrium moisture content is a simple and fast method for proper estimation of energy required for dehydration of maize and simulation of dried maize storage. Artificial neural network and thermodynamic equations for computation of maize heat and entropy of sorption were used, as a new method. The artificial neural network method for prediction of the equilibrium moisture content of maize was utilized. The heat of sorption of maize is predicted by a power model. After well training of equilibrium moisture content data sets using the artificial neural network models, predictive power of the model was found to be high (R<sup>2</sup> = 0.99). A power regression model was also developed for entropy of sorption. At moisture content above 11% (d.b.) the heat and entropy of sorption of maize decreased smoothly and they were highest at moisture content about 8% (d.b.).