Experimental and ANN modeling study on microwave dried onion slices

Abstract

The present work deals mainly with dehydration characteristics of onion slices. Microwave power levels of 100, 350, 550 and 750 W was practiced to dry onion slices with thicknesses of 2.5, 5, 7.5 and 10 mm. The results showed that moisture diffusivity and specific energy consumption of the process increased with both increasing microwave power and the samples thickness, and ranged from 0.82 × 10−8 to 6.13 × 10−8 m2 s−1 and from 0.82 to 5.43 MJ kg−1 water, respectively. The average activation energy varied in the range of 1.28–1.77. Furthermore, for simulation of drying process and to predict the moisture removal behavior of the samples, multi-layer feed-forward (MLF) artificial neural network (ANN) was employed. Practicing different networks and based on statistical parameters, the best topology, transfer functions and training algorithms were determined. The results revealed that, as a powerful tool, ANN modeling could be effectively used to predict drying kinetics and determine the moisture content of the samples.