烤烟主要物理指标的近红外模型研究

Studies on Near Infrared Model for Predicting Major Physical Indices of Flue-cured Tobacco

为了用近红外光谱分析技术快速测定烤烟的主要物理指标(叶面密度、填充值、厚度、平衡含水率和抗张力),采用近红外漫反射光谱法(NIDRS)和化学计量学方法建立了烤烟主要物理指标的近红外数学模型。结果表明,烤烟叶面密度、厚度、抗张力、填充值和平衡含水率模型预测均方差(RMSEP)分别为0.553,0.0169,0.149,0.234和0.277,平均相对偏差分别为7.82%,8.36%,10.88%,4.18%和2.83%,预测值和标准值之间没有显著性差异,可用于批量样品的快速检测。

In order to rapidly detect the major physical indices （weight per unit leaf area, filling value, thickness, equilibrium moisture content and tensile strength） of flue-cured tobacco with near infrared （NIR） spectroscopy technology, NIR mathematical models were developed with near infrared diffuse reflectance spectroscopy and stechiometry. The results showed that the root mean square errors of prediction （RMSEP） of weight per unit leaf area, thickness, tensile strength, filling value and equilibrium moisture content were 0.553, 0.0169, 0.149, 0.234, 0.277, respectively; their average relative deviation were 7.82%, 8.36%, 10.88%, 4.18%, 2.83%, respectively. The results of t-test indicated that there was no significant difference between predictive values and standard values, the developed NIR models could be used in the rapid detection of physical indices of batches of tobacco samples.