Note: Application of neural network modelling for the control of dewatering and impregnation soaking process (osmotic dehydration)

En este trabajo se ha elaborado un modelo de simulacion de transferencia de masa (p6rdida de agua y ganancia de sólidos) durante la deshidratacion osm6tica mediante un sistema de red neural (Ileuml Iletwork). Se emplearon dos redes por separado,una para la p6rdida de agua y otra para la ganancia de s6lidos. La validez del modelo se comprob6 con datos publicados previamente de gel de agar sumergido en una disoluci6n de sacarosa con un amplio intervalo de condiciones experimentales (temperatura de 30 a 70 °C, concentraci6n de sacarosa de 30 a 70 g/100, tiempo de 0 a 500 min y contenido de agar en el gel de 2 a 8%). Los mejores resultados se obtuvieron con 3 neuronas ocultas (hidden neurons), con lo que se consiguió predecir una transferencia de masa con una precision al menos tan buena como el error experimental para todo el intervalo de condiciones estudiadas. El inter6s tecnol6gico del modelo esta vinculado a la rapidez en la simulaci6n, comparable a una funci6n de transferencia tradicional, limitado numero de parametros y datos experimentales y del hecho de que no se necesitan consideraciones preliminares sobre los mecanismos propuestos.

The aim of this work was to elaborate a predictive model of the mass transfer (water loss and solute gain) that occurs during dewatering and soaking by using neural network modelling. Two separate feedforward networks with one hidden layer were used (for water loss and solute gain respectively). Model validation was carried out on results obtained previously, which dealt with agar gel soaked in sucrose solution over a wide experimental range (temperature, 30-70 °C; solution concentration, 30-70 g sucrose/100 g solution; time 0-500 min; agar concentration, 2-8%). The best results were obtained with three hidden neurons, which made it possible to predict mass transfer, with an accuracy at least as good as the experimental error, over the whole experimental range. The technological interest of such a model is related to a rapidity in simulation comparable to that of a traditional transfer function, a limited number of parameters and experimental data, and the fact that no preliminary assumption on the underlying mechanisms was needed.