A method for pH food dependent shelf‐life prediction of intelligent sustainable packaging device using artificial neural networks
Isadora Fernandez Brazolin, Felipe Matheus Mota Sousa, Jackson Wesley Silva dos Santos, Víktor Oswaldo Cárdenas Concha, Flávio Vasconcelos da Silva, Cristiana Maria Pedroso Yoshida
Abstract
Abstract Intelligent packaging systems can contribute to easily informing the quality of products. Tools such as artificial neural networks (ANN) have the potential to be used in the development of intelligent packaging. The objective of this work was the development and training of an ANN to facilitate the control of foods from the pH variation based on a sustainable colorimetric indicator of chitosan‐anthocyanin. The pH intelligent films were prepared with different concentrations of chitosan (Cch, 0.5%, 1.0%, and 2.0%, w/w) and anthocyanin (Cath, 0.5%, 1.0%, and 2.0%, w/w). The films were characterized by water solubility, mechanical properties, and thermal analyses. The colorimetric efficiency of the pH intelligent films was measured by immersing the material in a wide pH range (1.20–12.58) buffer solution, determining the color parameters L*, a*, and b*. From the experimental results, a database was built to develop an empirical multivariable model based on ANN. Higher Cch increases the solubility and resistance of intelligent films. Color variation was better identified in films containing Cath = 0.5%. The ANN presented assertiveness of 79%, showing that classification algorithms based on colorimetric measurements can be exploited to indicate alterations in food products resulting from pH variation.