Modellingcodfish drying: Comparison between artificial neural network, diffusive and semi-empirical models

Abstract

Convective drying is of prime importance in the food conservation industry and has beenconstantly studied and improved to obtain products with higher quality and lower processing time. In this work, three different models were usedto perform the codfish drying simulation: Artificial neural network (ANN), diffusive and semi-empirical models. The simulation results were compared for thefollowing experimental conditions: Drying air temperature of 20°C, air velocities of 2 and 3m/s and drying air relative humidities comprise between 55 and 65%. The simulations showed good results for the semi-empirical and ANN models, requiring improvements tothe diffusion model.