Quantitative determination method for middle flavor type of flue-cured tobacco leaves

Abstract

The invention discloses a method for conducting quantitative determination on a middle flavor type of flue-cured tobacco leaves on the basis of the detection contents of 40 types of substances. The method comprises the steps that the contents of the 40 substances in the tobacco leaves are determined according to tobacco industry standards and a literature method, the contents, obtained through detection, of the 40 substances are substituted into three classification functions as shown in the description respectively to calculate values of a Y freshness flavor, a Y strong flavor and a Y middle flavor, the three values are compared, and when the Y middle flavor value is the maximum, it is determined that the flavor type of the detected flue-cured tobacco leaves is the middle flavor type. According to the method for conducting quantitative determination on the middle flavor type of the flue-cured tobacco leaves, the determination accuracy of the middle flavor tobacco leaves is 94% or above, and a favorable technical support is provided for maintaining the stability of the sensory quality of cigarettes.

Description

technical field [0001] The invention belongs to the field of quality evaluation and control of flue-cured tobacco leaf aroma, and specifically relates to a method for quantitatively determining the intermediate aroma and aroma of flue-cured tobacco leaves based on the detection content of 40 different types of substances and combining the classification function coefficients to construct a Fisher classification function model Methods. Background technique [0002] According to the aroma style produced when it is burned, flue-cured tobacco leaves can be divided into three categories: light-flavored, middle-flavored and strong-flavored. The flavor type of flue-cured tobacco leaves largely determines the sensory style characteristics of different cigarettes, is an important factor to maintain the stability of the sensory quality of cigarettes, and is also one of the elements for industrial production and formula selection. At present, the evaluation of flue-cured tobacco flavo...

AI Technical Summary

Problems solved by technology

However, the substances used for regression judgment modeling are only aromatizing components, lacking comprehensive support for other types of substances, which is suspected of being partial and comprehensive. At the same time, more than 100 types of aromatizing components are used, and there are aromatizing component detection Problems such as slow speed, cumbersome qualitative retrieval, and difficult quantification; in addition, too many modeling variables will also lead to difficulties in regression modeling, so that the model is overfitted, resulting in low or unreliable prediction results.

Images