Abstract
To improve food production
via
fermentation with co-cultures of microorganisms (e.g., multiple lactic acid bacteria-LAB strains), one must fully understand their metabolism and interaction patterns in various conditions. For example, LAB can bring added quality to bread by releasing several bioactive compounds when adding soy flour to wheat flour, thus revealing the great potential for functional food development. In the present work, the fermentation of three soy and wheat flour mixtures is studied using single cultures and co-cultures of
Lactobacillus plantarum
and
Lactobacillus casei
. Bio-chemical processes often require a significant amount of time to obtain the optimal amount of final product; creating a mathematical model can gain important information and aids in the optimization of the process. Consequently, mathematical modeling is used to optimize the fermentation process by following these LAB’s growth kinetics and viability. The present work uses both multiple regression and artificial neural networks (ANN) to obtain the necessary mathematical model, useful in both prediction and process optimization. The main objective is to find a model with optimal performances, evaluated using an ANOVA test. To validate each obtained model, the simulation results are compared with the experimental data.