Method for searching analog tobacco leaf based on tobacco leaf near infrared spectra

Abstract

The present invention relates to a similar tobacco leaves search method based on the near infrared spectrum of tobacco leaves. The near infrared spectrum of tobacco leaves is used as basic data by the present invention. Distributed sampling is first carried out to the target tobacco leaves of each species; samples are pre-treated; the near infrared spectrum of the samples is obtained by scanning the samples on a near infrared spectrometer; principal component analysis (PCA) operation is carried out to a plurality of near infrared spectra of the target tobacco leaves of each species, obtaining loading matrixes, characteristic values and standardized residual errors, so as to generate a data model of the target tobacco leaves of each species; the near infrared spectrum of an unknown tobacco leaf, and the loading matrixes in the target tobacco leaf data models are used to carry out principal component decomposition calculation to the near infrared spectrum of the unknown tobacco leaf, so as to obtain the principal component score and decomposition residual error of the unknown tobacco leaf; the principal component score of the unknown tobacco leaf and the principal component space distance of the target tobacco leaf data models are calculated, and the residual error distance between the decomposition residual error of the unknown tobacco leaf and the standardized residual errors in the target tobacco leaf data models is also calculated; the distance between the unknown tobacco leaf and the target tobacco leaves is measured through the sum square root of the principal component space distance and the residual errors; the smaller the distance is, the higher the similarity is; finally, the distances between the unknown tobacco leaf and each target tobacco leaf is compared and sorted according to the size of the distances, so as to obtain a similar tobacco leaves search result.

Description

technical field [0001] The invention relates to a method for searching similar tobacco leaves in the tobacco industry. The search result is obtained by calculating the similarity between unknown tobacco leaves and various target tobacco leaves. Background technique [0002] The internal quality of tobacco leaves is affected by factors such as climatic conditions, planting regions, planting management, and processing. Tobacco leaf raw materials of various origins and varieties are mainly classified into different grades according to their appearance quality, and the correspondence between the appearance quality and internal quality of tobacco leaves is not exact. Therefore, before the tobacco raw materials are used in formula design and industrial production by tobacco processing enterprises, they need to be manually evaluated to determine their inherent quality characteristics. The sources of tobacco leaves involved in product design are complicated, and there are hundreds o...

AI Technical Summary

Problems solved by technology

In other industries, such as traditional Chinese medicine, the liquid chromatographic fingerprints of medicinal material extracts have been used for quality determination. However, the internal quality factors of tobacco have a w

Images