近红外分析技术在烟草中的应用  

The application of near-infrared spectroscopy in tobacco

作　　者：张翼[1] 杨征宇[1] 葛炯[1] 刘太昂[2] 张浩博[1] 蔡峰[1] 束茹欣 

机构地区：[1]上海烟草集团有限责任公司技术研发中心,上海200082 [2]上海大学材料科学与工程学院,上海200444

出　　处：《计算机与应用化学》2016年第2期251-254,共4页Computers and Applied Chemistry

基　　金：数字化辅助设计算法集成与应用技术研究(K2014-1-002z)

摘　　要：简要介绍近红外分析技术及其在烟草应用中需要的预处理、降维及建模方法,并总结了烟叶近红外光谱数据建立定性、定量模型过程的一般步骤。在此基础上结合我们的研究工作,重点介绍了近红外分析技术在烟叶、香烟及烟气中的定性判别及烟叶中化学成分的定量预报等应用。最后提出了关于近红外分析技术在烟草中应用需要深入研究的几个问题。Near-infrared spectroscopy （NIR） techniques and its applications in tobacco analysis were introduced in this paper. The techniques included the pretreatment, feature reduction and model construction of tobacco samples. At the same time, both qualitative and quantitative methods used widely in modeling were summarized. The successful applications of NIR techniques covered qualitative models for identifying tobacco leaves, cigarette brand and cigarette smoke as well as quantitative methods for predicting certain chemical substances in tobacco leaves Lastly, a few questions to the application of NIR techniques worthy for further researches on analyzing tobacco leaves were proposed.

关 键 词：近红外分析技术 支持向量机 主成分分析 烟草 

分 类 号：TQ015.9[化学工程] TP391.9[自动化与计算机技术—计算机应用技术]