Modeling and optimization of osmo-sonicated dehydration of garlic slices in a novel infrared dryer using artificial neural network and response surface methodology

摘要

This investigation aims to develop a novel infrared drying system and optimization ofultrasound-assisted osmotic dehydration of garlic slices. RSM and ANN approachesare used for modeling and optimization of the process parameters. The factors sonicationtime, sonication temperature, and osmotic concentration, were considered tooptimize responses, that is, solid gain (SG), weight loss (WL), rehydration ratio (RR),drying rate (DR), and allicin content (AC). The ultrasonic time and osmotic concentrationhad significant effects on mass transfer parameters and the quality of dried garlicslices. The optimum drying conditions were found at an ultrasonic temperature:20℃, ultrasound time: 28.54 min, and osmotic concentration: 55.58 while the optimumresponses were WL: 25.839, SG: 3.557, RR: 7.512, DR: 0.163 (g H_2O/g.dm/min), and AC: 15.229 (mg/g). The prediction of the responses revealed that bothRSM and ANN approaches can predict the model precisely but the ANN modelsshowed higher accuracy.