应用近红外光谱或化学指标判别不同产地晒红烟的相似性

近红外光谱采集物质信息较全面，是多种物质的复合信息，但具体特征不明显；化学指标虽能够体现物质的具体特征，但所包含的物质信息却不够全面。以2012年和2013年贵州、湖南、吉林、江西、山东、四川六个产地共115个晒红烟样品为研究对象，其近红外光谱采用一阶导数和平滑处理，测定的26个化学指标及其26个计算值(譬如糖碱比等)采用归一化处理，应用PPF投影技术对不同产地晒红烟进行相似性分析。结合近红外光谱和化学指标的分析结果表明：应用近红外光谱和化学指标分析六个产地晒红烟分布规律较为一致，表明通过近红外光谱信息和大量检测数据均可分析产地特征的相似性；对52个化学指标的方差贡献率分析得出，决定产地特征的重要指标是亚硝胺；应用近红外光谱对小产地之间进行聚类分析，得出部分小产地之间具有可替代性，结合主要内在的化学指标工业需求取向，可合理替代不同烟叶原料，降低产品加工质量的波动性。

Analyzing the Similarity of Different Growing Areas in Red Sun-Cured Tobacco by Using Near Infrared Spectroscopy or Chemical Data

Near infrared spectroscopy is a comprehensive information collection material, composite information of various substances, but the specific feature is not obvious; the chemical index can reflect the specific characteristics of the material, but the material information is not comprehensive enough. In this paper, total of 115 red sun-cured tobacco samples in 2012 and 2013 were tested, which were from Guizhou, Hunan, Jilin, Jiangxi, Shandong, Sichuan. After near infrared spectroscopy was processed by the first derivative and SG smooth and 26 items of chemical data and 26 items of calculated data (for example, ratio of sugar to nicotine, etc.) were normalized, the PPF (Projection of Basing on Principal Component and Fisher Criterion) projection method was used to analyze the similarity and substitution of growing areas in red sun-cured tobacco. Combined with the analysis of near infrared spectroscopy and chemical indicators, the results showed that the similarity analysis result by using near infrared spectroscopy is in basic agreement with the result by using chemical data, which means two methods can both analyze the similarity of growing areas. Nitrosamines would determine the growing areas as an important indicator by analyzing the variance contribution rate of 52 test data. The similarity of small areas was analyzed by used near infrared spectroscopy and the result was that the part of different small areas can be replaced. Application of near infrared spectroscopy technology can quickly and accurately analyze the similarity and substitution of different growing areas. Combining chemical data can be used to analyze its intrinsic characteristics, which can play an important role in the scientific exploitation of tobacco raw materials.