Effect of temperature, starch and plasticizer concentrations on color parameters of ulluco (Ullucus tuberosus Caldas) edible films

3rd International Congress of Mechanical Engineering and Agricultural Science

机械制造;农业科学

Homez, A.K.^1 ; Daza, L.D.^1 ; Solanilla, J.F.^2 ; Váquiro, H.A.^1

Centro de Desarrollo Agroindustrial Del Tolima, CEDAGRITOL, Departamento de Producción y Sanidad Vegetal, Facultad Ingeniería Agronómica, Universidad Del Tolima, Ibagué

730006, Colombia^1

Departamento de Agroindustria, Facultad de Ciencias Agrarias, Universidad Del Cauca, Popayán

190003, Colombia^2

关键词: Bayesian regularization;    Complex relationships;    Effect of temperature;    Feed-forward architectures;    Glycerol concentration;    Sensibility analysis;    Starch concentration;    Training algorithms;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/437/1/012003/pdf
DOI  :  10.1088/1757-899X/437/1/012003

学科分类：农业科学（综合）

来源: IOP

【 摘 要 】

The growing concern about limited natural sources and pollution has aroused the interest of developing bio-packaging based on agricultural polymers. The Andean tuber ulluco (Ullucus tuberosus Caldas) is a promising source of starch that can be used in the production of edible films or coatings for preserving the quality of food products. Color is an important feature for edible films production due to can alter the visual perception of the coated product. In this study, the effect of drying temperature (T) (40 to 60°C), starch concentration (SC) (2.0 to 3.0% w/v) and glycerol concentration (GC) (1.0 to 3.0% v/v) on the color parameters (L∗, a∗ and b∗) of ulluco edible films was evaluated. A composite central design was used to combine five levels of the three factors (T, SC and GC). Artificial neural network (ANN) were applied to establish mathematical models that could represent the dependence of the factors on the color parameters. The edible films showed luminosity values between 87.4 and 90.3. Results indicated that the higher T, the lower a∗ and b∗ parameters and the higher luminosity of the films. Likewise, the lower G, the higher luminosity of edible films. Sensibility analysis shows that SC has an important influence on L∗, a∗, b∗ parameters. Statistical results indicated that the ANN approach provided an accurate estimation (R 2 adj = 0.885 and MSE = 0.054) of the color parameters. ANN use a feed-forward architecture with five neurons in the hidden layer, hyperbolic-tangent transfer function, and Bayesian regularization backpropagation as training algorithm. ANN could be an alternative tool for modelling the complex relationship between the color parameters and the temperature, starch and plasticizer concentrations on ulluco edible films.