Application of Artificial Neural Network for Monitoring the Pre Ultrasound-Osmosed Carrot Air Drying

چکیده مقاله

 This thesis objective is to utilize image analysis and artificial neural network to predict mass transfer kinetics and color changes of osmotically dehydrated assisted ultrasound carrot slices. Samples were osmotically dehydrated through three different solutions (Nacl, Sorbitol and Nacl/Sorbitol 20 % w/w), 5, 25 and 50ºc, 7 hours osmotic duration (60, 120, 180, 240, 300, 360, 420 min) and ultrasound treatment (5 min) in three temperature(5, 25 and 50ºc). Artificial neural network with three activation function (hyberbolic tangent- hyberbolic tangent, sigmoid logarithm- sigmoid logarithm and hyberbolic tangent- sigmoid logarithm), were developed. Time, temperature and final moisture content was used as inputs and water loss, solid gain and ∆E was used as outputs layers. It was found that artificial neural network with 1 to 30 neurons in two layers gives the best fitting, which made it possible to predict water loss, solid gain and color change, with percentage of relative mean squre error (%MRE) 0.024, 0.018 and 0.050 respectively and R2 0.969, 0.969 and 0.904 respectively.

دریافت لینک دانلود مقاله