Pit mud quality detection method based on near infrared-intermediate infrared fusion algorithm

Abstract

The invention discloses a pit mud quality detection method based on a near infrared-intermediate infrared fusion algorithm, which comprises the following steps: 1, obtaining wine pit mud samples with different qualities, measuring the calcium content, and collecting near infrared spectrum and intermediate infrared spectrum data for data fusion; 2, feature variables are extracted by using an uninformationized variable elimination algorithm UVE, a competitive adaptive reweighted sampling algorithm CARS and a competitive adaptive reweighted-uninformationized variable elimination sampling algorithm CARS-UVE, and the extracted feature variables are fused again by using a near infrared-intermediate infrared fusion algorithm; and 3, establishing a pit mud calcium quantitative model for realizing pit mud quality detection by using the fused characteristic variables and calcium content data. According to the method, the quality of the cellar mud can be evaluated more scientifically, quickly, economically and environmentally, so that the judgment of a workshop on the quality of the wine cellar mud does not stay on sensory identification, and the accuracy and effectiveness of a judgment result can be improved.

Description

technical field [0001] The invention belongs to the analysis technology of wine cellar mud quality in the food field, and belongs to the technical field of analysis and detection. Background technique [0002] Cellar mud plays an important role in the fermentation of Luzhou-flavor liquor. Pit mud is continuously domesticated and matured through artificial production, fermentation culture and long-term continuous production. During this period, the physical and chemical properties of the pit mud continued to change, and finally became a high-quality old pit mud that can produce high-quality liquor. Therefore, in liquor production, it is necessary to monitor the quality change of pit mud. [0003] The traditional method of judging the quality of pit mud is through sensory analysis, which is highly subjective. Since pit mud is a microbial enrichment, in recent years, the quality of pit mud has also been judged by the number and types of microbial colonies in pit mud. Howeve...

AI Technical Summary

Problems solved by technology

The accuracy of the model established by the elimination of non-informative variables and the competitive adaptive reweighting algorithm is good, but the number of selected variables is large; if the two are used together, the elimination of non-informative variables-competitive adaptive reweighting is formed Algorithm, the number of wavenumber variables is reduced, but the model is prone to overfitting

Single or joint feature variable selection algorithms cannot take into account the accuracy of the model and the number of variables, and there are problems such as low prediction accuracy, complex models, and low detection efficiency, which are limited in practicability

Images