Study on Modeling Method of Chemical Composition of Tobacco for Micro Near Infrared Instrument

The quality analysis of the chemical composition in the process of tobacco purchasing and baking has become focus of the enterprises attention. At present, the chemical composition analysis in conventional laboratories is time-consuming and laborious, which cannot meet the practical application requirements. There is an urgent need for the instrument with low development cost and miniaturization for tobacco quality analysis. Therefore, based on DLP near infrared analysis module DLPNIRNANOEVM of TI company, this paper studies the modeling method of tobacco chemical composition. The spectral data of samples collected by micro near infrared instrument has the characteristics of low resolution, high noise and weak signal. Therefore, the basic PLS modeling method and spectral preprocessing method are optimized in this paper, and the mathematical models of total sugar, reducing sugar, total nitrogen and nicotine in tobacco are established according to the relationship between content values and the spectra. The experimental results show that this method can effectively improve the accuracy and robustness of the model, which provides an important reference for the practical application of micro near infrared spectrometer.