A processing method and processing system for sensory evaluation and smoking data of tobacco leaf products

Abstract

The invention discloses a processing method and system for sensory evaluation data of tobacco leaf products. The scheme obtains individual sensory evaluation data and uses Monte Carlo sampling to construct multiple training data subsets to calculate pairwise sensory evaluation data. The correlation of smoking indicators was improved to improve the representativeness of samples, and statistical methods were used to test abnormal sampling and inversely deduce the abnormal samples of sensory evaluation data. The data evaluation and smoking method based on the correlation between the sensory evaluation and smoking evaluation indicators proposed by the present invention evaluates the reliability of the smoking evaluation data, finds out abnormal data in the smoking evaluation data and eliminates them, and improves the accuracy and effectiveness of the smoking evaluation data of tobacco products. In order to improve the efficiency of the use of evaluation data and reduce the workload of evaluation personnel.

Description

technical field [0001] The invention belongs to the field of quality detection and quality control of tobacco leaf products, and in particular relates to a method for judging the quality stability of tobacco leaf products through sensory evaluation and smoking. Background technique [0002] The national cigarette standard "GB T 5606.6-2005 Cigarette Part 6: Comprehensive Judgment of Quality" stipulates that the sensory evaluation of smoking quality is a mandatory technical requirement: if there is a class A quality defect in the sensory quality, the batch of cigarettes will be judged as unqualified. Therefore, the quality of sensory evaluation smoking is an important indicator for tobacco workers to evaluate tobacco products, and the reliability of sensory evaluation data will directly affect the final quality of tobacco products. [0003] Today, the evaluation of the quality of tobacco leaf products is to select a certain number of professionals with rich experience and div...

AI Technical Summary

Problems solved by technology

However, in the actual process of sensory evaluation and smoking, it is impossible to evaluate and smoke each tobacco leaf sample multiple times from the beginning to the end. On the one hand, because the workload of smoking evaluation is too large, it is very easy to bring its own sensory fatigue and cause deviations in the evaluation data; On the other hand, because the selection of multiple evaluations involves the standardization of samples, the standardization of sampling and sample preparation is likely to cause random deviations.

Images