摘要
CO,焦油,烟气烟碱是卷烟产品的重要指标,传统方法测量时,有污染并且不能预测。根据卷烟产品粉末的近红外光谱,用30个品牌样品,600条光谱建立了焦油、CO、烟碱三大烟气指标的校正模型,用于准确测定和预测卷烟产品的三大烟气指标。在所有的校正模型中,原始谱图经过二阶导数和偏最小二乘(PLS)处理,实验结果的偏差比较小,表明近红外光谱技术在分析卷烟产品的化学指标时可以代替经典化学方法,可以为烟草行业节省大量资金和显著提高工作效率。另外还利用近红外光谱图对卷烟产品的价格档次进行预测分类,结果令人满意,可以对卷烟产品价格的制定提供参考。
CO,tar,nicotine are important indexes of cigarette products.There is pollution and they can not be predicted when the traditional method is usedd to measure the indexes.The calibration models of tar,CO,nicotine indexes were established according to 30 brand samples,600 spectra and the near-infrared spectrum of cigarette powder to accuratly measure and predict the three flue gas indexes of cigarette products.The original spectrograms were treated by second-order derivative and partial least square（PLS） during the correction of all the models.Deviation of experimental results is relatively small.This shows that the near-infrared spectroscopy can replace classical chemical methods in the chemical indexes analysis of cigarette products,and save a lot of money and improve the work efficiency for the tobacco industry significantly.In addition,the article also predicted and classified the price levels of cigarette by the near-infrared spectrograms.It can provide a reference for the formulation of the cigarette price.
出处
《现代电子技术》
2012年第16期101-103,共3页
Modern Electronics Technique
