论文部分内容阅读
以高效液相色谱(HPLC)分析结果为参考值,建立了快速测量茶多酚中总儿茶素含量的近红外光谱定标模型.将48份茶多酚样品组成定标样品集,在1000～2500nm(4000～10000cm-1)的近红外漫反射光谱为定标波长范围内,光谱经一阶导数(Firstderivative)、二阶导数(Secondderivative)、标准归一化(Stan-dardnormalvariate,SNV)和多元散射校正(multiplicativesignalcorrection,MSC)处理后结合偏最小二乘回归(PLS)定标.经内部交叉验证表明,光谱经SNV处理后建模结果最佳.模型的相关系数Corr.Coeff=0.997,校正均方根RMSEC=1.71%.比较了经典最小二乘法(CLS)、偏最小二乘法(PLS)和主成分回归(PCR)等方法建模结果,以偏最小二乘回归建模效果最好.
A high-speed liquid chromatography (HPLC) analysis was used as a reference to establish a near-infrared calibration model for rapid measurement of total catechins in tea polyphenols. 48 tea polyphenol samples were used to set the calibration sample set. The near-infrared diffuse reflectance spectrum of ~2500 nm (4000 to 10000 cm-1) is within the calibration wavelength range, and the spectrum is obtained by first derivative, second derivative, standard normalization (Stan-dard normal variation, SNV) and Multivariate scattering correction (MSC) was combined with PLS calibration. Internal cross-validation showed that the results obtained after SNV treatment were the best. The correlation coefficient of the model was Corr.Coeff=0.997. Root Mean Square RMSEC = 1.71%. The modeling results of classic least squares (CLS), partial least squares (PLS), and principal component regression (PCR) methods are compared. The modeling results are better with partial least-square regression.