Premium
Modeling of vacuum‐ and ultrasound‐assisted osmodehydration of carrot cubes followed by combined infrared and spouted bed drying using artificial neural network and regression models
Author(s) -
Alizehi Mohammad Hashem,
Niakousari Mehrdad,
Fazaeli Mahboubeh,
Iraji Maryam
Publication year - 2020
Publication title -
journal of food process engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.507
H-Index - 45
eISSN - 1745-4530
pISSN - 0145-8876
DOI - 10.1111/jfpe.13563
Subject(s) - osmotic dehydration , artificial neural network , vacuum drying , chemistry , linear regression , mean squared error , partial least squares regression , organoleptic , analytical chemistry (journal) , materials science , food science , mathematics , chromatography , freeze drying , mass transfer , machine learning , statistics , computer science
This research seeks to understand the effects of using various vacuum pretreatments (0 and 200 mBar) and different ultrasound power levels (0 and 130 W) on osmodehydration and spouted bed drying of carrot cubes. Effects of using diverse air‐drying temperatures (55, 65 and 75°C), infrared power (0 and 250 W) and the presence of Teflon beads as inert particles (0 and 30% w/w), were also investigated. The effects of above‐mentioned parameters on drying time and physicochemical characteristics of dried carrot samples including hardness, shrinkage, total carotenoid content, rehydration ratio and total color difference (∆E) were examined. The results explicitly demonstrated that the utilization of osmodehydration in conjunction with ultrasound and vacuum (UVOD) had shorter drying time compared to those just treated with osmotic dehydration. The highest total carotenoid content and the lowest ∆E were belonged to the samples exposed to UVOD. The efficiency of artificial neural network (ANN), genetic algorithm based ANN (GANN) and Regression mathematical models were investigated to determine the nonlinear relationship between the inputs and outputs variables. To find the efficiency of all models, mean square error were compared. The results indicate that the ANN model could provide more reliable prediction over the other models. Practical application Combined IR‐convective drying has gained increased attention in the recent years because of the fast and efficient transfer of heat, cheap processing cost, and uniform heat transfer, which resulted in products with better organoleptic properties than common drying processes. The combination of hot air and IR radiation resulted in a synergistic effect, leading to more efficient drying than either individual convective or IR drying. Vegetables with high levels of moisture content can be effectively dried using pretreatments such as ultrasonically vacuum‐assisted osmotic dehydration, which reduced the drying time.