近红外光谱相似度匹配分析不同陈化条件下烟叶化学成分的差异性

Analysis Method of Similarity Match of NIR Spectral in Differences of Chemical Components of Tobacco Leaves under Different Aging Conditions

应用近红外光谱相似度匹配的定性分析功能和SPSS 18软件中的重复方差分析方法,研究4种橡木桶陈化烟叶与自然陈化的烟叶的化学成分在一年陈化时间中差异性的变化。通过研究表明:不同的陈化条件下,4种橡木桶陈化烟叶的化学成分都发生了一定的改变,与自然陈化烟叶的相似度匹配值都低于90,且相互间的差值在10左右。通过对TQ统计分析软件建立的模型结果的检验证明,模型的计算准确率高,可达100%,能有效区分不同陈化条件下烟叶的化学成分差异,可以作为卷烟生产企业寻找最佳陈化条件的一种补充手段。

Two methods were suggested to analysis the differences of chemical components of five kinds of tobacco leaves under different aging conditions in one year. In this research, similarity math identifying models was established by using the TQ statistical software and the data of NIR diffuse re- flection spectral of four oak samples. Then, the repeated measurement ANOVA of SPSS 18 software was used to study the differences. The results showed that the different aging conditions could ehange the chemical components of tobacco leaves. Four tobacco similarity matching values were less than 90 and the difference between each other was 10. Therefore, the application of FT-NIR spectroscopy could identify the differences of chemical components of tobacco leaves under different aging condi- tions and its accuracy rate was up to 100%. It provided a new method to find the best aging condi- tions.