A PROPOSAL FOR ISOTHERMAL MODEL SELECTION, A HEURISTIC APPROACH Cover Image

A PROPOSAL FOR ISOTHERMAL MODEL SELECTION, A HEURISTIC APPROACH
A PROPOSAL FOR ISOTHERMAL MODEL SELECTION, A HEURISTIC APPROACH

Author(s): Dragos Arotaritei, Andrei Gheorghita, Mihai Ilea, Carmen Lucia Trinca
Subject(s): Methodology and research technology, ICT Information and Communications Technologies
Published by: Carol I National Defence University Publishing House
Keywords: adsorption isotherm; mathematical models; decision systems; goodness of fit models;

Summary/Abstract: There are many mathematical models for sorption isotherms, with some practically application in the food chemistry. In this case, an isotherm describes a relationship (linear or nonlinear) between moisture content and water activity in a certain type of food. The hysteresis for a typical food product is taken into account with educational objective: a better understanding of the sorption process. Depending of the food chemical composition, the relationship between water activity and water content can be simple of very complex, a large number of models having been proposed to cover a large variety of curves. There is no a clearly classification of types of isotherms but using the literature a structure based of four classes of models is proposed to user for selection: absorption empirical models, models based on Polanyi’s theory, chemical adsorption isotherms and physical adsorption isotherms. The relationship between water activity and food moisture content is complex, depending on the food chemical composition; there is a particularly class of isotherms. A good choice is a difficult task and furthermore, some of the parameters could be inadequate for some experimental data related to some type of food based only on the analysis of shape. The user will have the opportunity to do the agreement among possible solution for the problem given by a set of models according to physical meaning of the process. The finding of the most fitting parameters of the model, no matter of the method depends on the initial condition from where starts the search algorithm. A combination of statistical methods, linearization and heuristics are used to find the most suitable initial solution in order to have a convergence of to optimal parameters that fit best the experimental data. In order to increase the generality, the didactical tool includes the most known models of sorption that is, the processes of adsorption, desorption and absorption in general.

  • Issue Year: 17/2021
  • Issue No: 03
  • Page Range: 256-261
  • Page Count: 6
  • Language: English
Toggle Accessibility Mode