The objective of this research was to predict drying behavior of hot air drying using an empirical model (EM) and an artificial neural network model (ANN). Rubber sheet with initial moisture content ranging of 23-40% dry-basis was dried by temperature ranging of 40-70°C and air flow rate of 0.7 m/s. The desired final moisture content was set at 0.15% dry-basis. The results showed that drying rate of rubber sheet dried with hot air convection was faster than conventional natural aeration. The EM and ANN were simulated to describe the drying behavior of products. Furthermore, prediction results between EM and ANN were compared with the experimental data. In this research, it was obviously found that ANN can describe the drying behavior effectively. Additionally, it was also found that predicted results of Multilayer feed forward Levenberg-Maqurdt’s Back-propagation ANN were good agreement with the experimental results compared to those results of EM. It is the optimum architecture for prediction the evolution of moisture transfer for hot air drying.
Hot air drying characteristics of mango ginger: Prediction of drying kinetics by mathematical modeling and artificial neural network