Rapid monitoring of beer-quality attributes based on UV-Vis spectral data

Abstract

This work aimed to determinate eight beer properties using UV-Vis spectra in combination with principal component regression (PCR) or artificial neural network (ANN) models. A statistical experimental design was performed to generate the calibration data. First, principal component analysis (PCA) was applied to the original spectral data, and the scores in significant PCs were utilized to calibrate both models. PCR showed poor correlation for beer parameters (R 2 < 0.61). The ANNs showed satisfactory correlations (R 2 = 0.74-0.92) and low relative error considering a variable range (E r < 9%) for most of the beer-quality attributes, but vicinal diketones (R 2 = 0.56, E r = 16.69%). Once implemented, this method would be fast and low cost. ARTICLE HISTORY

Brewing process optimization by artificial neural network and evolutionary algorithm approach

“…The primary data utilized in this work, related to beer samples characterization, were previously reported by senior authors in a complementary manuscript dealing with brewery process monitoring (Coelho de Oliveira et al, ). The ranges identified in quality parameters of the beer samples were typical for Ale‐type beers (Briggs et al, ).…”

“…Besides, the ranges for all beer quality parameters were broad enough to ensure calibration of RSM and ANN models. The variability of quality attributes in the data set was a consequence of changes in beer ingredients and protease reaction time derived from the experimental design (Online Resource 1) (Coelho de Oliveira et al, ).…”

“…Beer production at bench scale was based on Ale‐type formula proposed by Palmer (), which was previously described (Coelho de Oliveira, Elias da Cunha Filho, Rocha, & Fernández Núñez, ). The beer substrate was composed by malted barley (Pilsen malt, Agrágria, Brazil), malted wheat (Light wheat malt, Weyermann, Germany) and high maltose corn powder (Cargill, Brazil) with a total mass of 160 g. Their proportions were defined by an experimental design, which is described in Section 2.2.…”

“…Seven parameters were analyzed under the characterization, all of them are briefly described below. They are included among the most frequent quality attributes used to characterize beers (Briggs, Boulton, Brookes, & Stevens, ; Coelho de Oliveira et al, ). These parameters are very related to customer perception and palatability. Foam stability : it was determined by Rudin's method using nitrogen instead of carbon dioxide (3 L/min) (Rudin, ).…”

Chemometrics in analytical chemistry – an overview of applications from 2014 to 2018

“…The ANN was implemented using the Levenberg-Marquardt training algorithm, being able to create more accurate models than PLS 74 . Oliveira et al 75 determined protein concentration, foam stability, haze, color, total acidity (TA), alcohol content, and bitterness in Ale beers properties using UV-Vis spectra in combination with PCR and also with ANN models.…”