Optimality in Microwave-Assisted Drying of Aloe Vera (Aloe barbadensis Miller) Gel using Response Surface Methodology and Artificial Neural Network Modeling

The present work illustrates the Microwave-Assisted Drying (MWAD) characteristic of aloe vera gel combined with process optimization and artificial neural network modeling. The influence of microwave power (160–480 W), gel quantity (4–8 g) and drying time (1–9 min) on the moisture ratio was investigated. The drying of aloe gel exhibited typical diffusion-controlled characteristics with a predominant interaction between input power and drying time. Falling rate period was observed for the entire MWAD of aloe gel. Face-centered Central Composite Design (FCCD) developed a regression model to evaluate their effects on moisture ratio. The optimal MWAD conditions were established as microwave power of 227.9 W, sample amount of 4.47 g and 5.78 min drying time corresponding to the moisture ratio of 0.15. A computer-stimulated Artificial Neural Network (ANN) model was generated for mapping between process variables and the desired response. ‘Levenberg–Marquardt Back Propagation’ algorithm with 3-5-1 architect gave the best prediction, and it showed a clear superiority over FCCD. © 2016, The Institution of Engineers (India).