论文部分内容阅读
目的:利用黄芩提取物样品的近红外漫反射光谱(NIRS)信息,建立能够快速分析其3种有效成分含量的校正模型。方法:共收集12个不同厂家的100批样品,其中80批样品作为校正集,20批样品作为验证集,结合偏最小二乘法(PLS),建立了黄芩提取物中黄芩苷、黄芩素和汉黄芩素3种有效成分的近红外定量校正模型。结果:3个校正模型的建模效果均较好,交叉检验决定系数(R2CV)分别为0.994 8,0.998 7,0.994 8,校正均方差(RMSEC)分别为0.440,0.022 5,0.011 1,交互验证均方差(RMSECV)分别为2.259,0.055 3,0.048 3。用验证样品进行外部验证,预测相关系数(r2)分别为0.998 2,0.996 5,0.990 9,预测均方差(RMSEP)分别为0.486,0.027 1,0.011 0。结论:结果表明,近红外光谱技术可对黄芩提取物中黄芩苷、黄芩素和汉黄芩素含量进行简便、快速、准确分析。
OBJECTIVE: To establish a calibration model capable of rapidly analyzing the content of three active ingredients of Scutellariae Radix by using near infrared diffuse reflectance spectroscopy (NIRS) information. Methods: A total of 100 batches of samples from 12 different factories were collected. Among them, 80 batches of samples were used as calibration sets and 20 batches of samples were used as validation sets. Combined with PLS, baicalin, baicalein and baicalein Near Infrared Quantitative Correction Model of Three Kinds of Active Components of Baicalein. Results: The modeling results of the three calibration models were good, with R2CV of 0.994 8,0.998 7,0.994 8 and RMSEC of 0.440, 0.022 5 and 0.011 1, respectively Mean square error (RMSECV) were 2.259,0.055 3,0.048 3 respectively. The validated samples were validated externally with correlation coefficients (r2) of 0.998, 2.0996 5 and 0.990 9 respectively, and the RMSEPs were 0.486, 0.027 and 1.01011, respectively. Conclusion: The results show that NIR spectroscopy can be used to analyze the content of baicalin, baicalein and wogonin in Scutellaria baicalensis Georgi easily, rapidly and accurately.