粒度对烟末总糖、总氮和烟碱含量NIR预测值的影响

Effects of Particle Size on Estimated Values of Total Sugar, Total Nitrogen and Nicotine Contents in Tobacco Fines with Near-infrared Spectroscopy

考察了粒度对烟末总糖、总氮和烟碱含量近红外(NIR)预测值的影响.结果表明,在20～100目范围内,随着粒度的降低,烟末总糖、总氮和烟碱含量的NIR预测值与其化学测定值的相对误差逐渐减小.当烟末粒度≥40目时,其预测误差＜5%,RSD＜8%.因此,应用近红外光谱分析技术预测烟末样品中总糖、总氮和烟碱含量时,样品粒度在40目以上才能保证预测数据的准确性和精密性.

The effects of particle sizes of tobacco fines on the estimated values of total sugar, total nitrogen and nicotine contents with NIR were studied. The results showed that, the relative errors between estimated and analytical values of total sugar, total nitrogen and nicotine reduced with the decrease of particle sizes in the range of 20-100 meshes. The estimated error was under 5%, and RSD under 8% when the particle size was≥40 meshes. The conclusion was that the estimate accuracy and precision of said chemical components would not be guaranteed unless the particle size of tobacco samples was over 40 meshes.