摘要
应用傅立叶变换近红外光谱法测定730个具有代表性的烟草近红外光谱数据,采用偏最小二乘回归建立了近红外光谱信息与其含量之间的定量校正模型,并对50个验证样品进行预测验证。总挥发碱和总挥发酸的预测标准差(RMSEP)分别为0.020和0.009,验证样品的相对标准偏差各为1.120%和0.919%。
The near infrared （NIR） spectra of 730 representative tobacco samples were measured with Fourier transform near infrared spectroscopy. The calibration models between spectral data and the contents of total volatile base （TVB） or total volatile acid （TVA） of tobacco samples were established by partial least square regression and verified with a set of 50 tobacco samples. The root mean square errors of prediction （RMSEP） for TVB and TVA were 0.020 and 0.009, respectively; the relative standard deviations of TVB and TVA were 1. 120% and 0.919% respectively for the 50 validation samples.
出处
《烟草科技》
EI
CAS
北大核心
2006年第3期33-37,共5页
Tobacco Science & Technology
基金
国家烟草专卖局资助项目(国烟科[2003]542号)"应用近红外检测技术快速测定烟叶主要化学成分(20项指标)的研究"的一部分。
关键词
近红外光谱法
烟草
总挥发酸
总挥发碱
Near infrared spectroscopy
Tobacco
Total volatile acid
Total volatile base
