烟草近红外光谱分析结果影响因素综述

Research Progress on Factors Affecting the Determination of Tobacco Constituent with Near Infrared Reflectance Spectroscopy

结合国内外研究现状,针对校正集样品选择,样品粒度、含水率、试验因素以及光谱预处理和数学建模方法的选择对烟草近红外光谱分析结果的影响进行了探讨分析。结果表明:为获得可靠的分析结果,样品选择应均匀、广泛,有较好的代表性,并考虑样品粒度、含水率及其他因素的影响,同时应选择合适的光谱处理方法和数学建模算法。

The analysis of tobacco constituent with near infrared spectroscopy was a rapid and accurate technique.The effect of sample handling,chemical analysis as well as data processing methods were investigated with regards to the performance of the near infrared spectroscopy in determination of tobacco constituents.The analysis parameters were： range of samples for calibration;water content and granularity of sample;error relating to chemical analysis;sample loading and test condition;pretreatment method for NIR spectrum;regression method for calibration.In order to obtain reliable results,it was advisable to take these variables into consideration when NIR model was established for the estimation of the ingredients of tobacco.