Flavor and Rapid Prediction of Red Wine by the Chemometrics Algorithm Based on Multidimensional Spectral Data.

Since its birth, red wine has been loved by people of all walks of life. The taste of red wine has changed and the pursuit of quality has always been the most sought-after goal by sommeliers, winemakers, and the public. However, due to the rich taste of red wine, any link is willing to produce different flavors. At present, there is no quantitative control study on the flavor of red wine. The purpose of this paper is to analyze the flavor of red wine through the chemometric algorithm and establish a reasonable model to predict the flavor of red wine. Aiming at the research of red wine flavor, this paper designs a red wine flavor extraction experiment and extracts the substances that produce an aroma and flavor in red wine to the greatest extent through strict selection of extraction head and reaction time. For the rapid analysis of red wine flavor, this paper quantitatively describes the chemical category, volatilization time, molecular weight, etc., of flavor substances by analyzing the multidimensional spectral data of red wine, so that flavor substances can be quickly located. The experimental results of this paper prove that, for different red wines, the algorithm in this paper can accurately identify the flavor substances in red wine. Also, for red wine multidimensional spectral data, the algorithm in this paper can improve the accuracy by 30% and save the running time by 30%. This shows that the research in this paper can analyze and quickly predict the flavor of red wine.