Calibration models for routine analysis of cider by mid-infrared spectroscopy

Abstract

Partial least square regression was used to develop different calibration data sets for multi-parameter routine analysis of ciders. Parameters included were those related with the monitoring of fermentation process and cider quality: specific gravity, total and volatile acidities, alcoholic proof, pH and fructose. Calibration performances were evaluated by means of the prediction residual sum of squares (PRESS), the root mean squared prediction error of cross-validation (RMSECV) and the residual predictive deviation (RPD) values (ratio of the standard deviation of the population to the standard error of cross-validation). Validation of the models was assessed in terms of accuracy and precision. Mean recoveries of the predicted results compared to the reference values were close to 100%, with repeatability and reproducibility similar to those of the reference methods.