The invention discloses a method for predicting the year of Maotai-
flavor liquor based on the combination of SPME-Arrow / GC-MS. The method comprises the following steps: adding
sodium chloride and an
internal standard into a to-be-detected
liquor sample, fully and uniformly mixing, carrying out on-
machine detection by using SPME-Arrow / GC-MS, collecting data, carrying out peak extraction, peak alignment and qualitative and quantitative identification analysis, and determining the year of the Maotai-
flavor liquor according to the year of the Maotai-
flavor liquor. Carrying out orthogonal partial
least squares-discrimination and
single factor variance analysis on the result
data set, screening by taking VIP greater than or equal to 1.1 as a standard under the condition that p is less than 0.05 to obtain characteristic compounds of the
white spirit in different years, and then calculating the characteristic compounds of the
white spirit in different years by taking the characteristic compounds as input variables of a
random forest model and taking different years of the
white spirit sample as target variables of the model. And training and testing the model to obtain a year prediction model, and finally inputting the characteristic compound of a to-be-tested
wine sample into the
random forest year prediction model, so that a prediction result output by the model is the year of the to-be-tested sample. The method is simple to operate, time-saving, labor-saving and high in detection sensitivity, and a new technical thought is provided for
vintage wine identification.