天舒胶囊乙醇提取过程的近红外光谱快速测定方法研究

Fast measurement method based on near infrared spectroscopy in extraction process of Tianshu capsules

基于近红外光谱法,采用偏最小二乘法(PLS)分别建立了天舒胶囊乙醇提取过程固含量和阿魏酸浓度的定量校正模型,结果显示,固含量校正集和交叉验证集模型的相关系数(R2)分别为0.967 301和0.947 726,校正集误差均方根(RMSEC)为0.054 7,交互验证集误差均方根(RMSECV)为0.069 8;阿魏酸校正集和交叉验证集模型的相关系数(R2)分别为0.986 879和0.962 243,RMSEC为1.402 6,RMSECV为2.400 2。采用所建模型进行在线分析,固含量和阿魏酸的预测结果与实测值的相关系数(R2)为0.993 3和0.991 6,预测误差均方根(RMSEP)分别为0.039 3和1.669 3,预测平均相对偏差(RSEP)分别为3.49%和3.58%。结果表明,该文所建模型可用于天舒胶囊乙醇提取过程固含量和阿魏酸浓度的快速测定。

Based on the near infrared spectroscopy,partial least square（ PLS） method was used to respectively develop the quantitative calibration models to fast measure the contents of the total solid and ferulic acid in extraction process of Tianshu capsule extracts.The results showed that in the quantitative model of solid content,the correlation coefficients（ R2） of calibration set and cross validation set were 0. 967 301 and 0. 947 726. The root-mean square error of calibration set（ RMSEC） was 0. 054 7 and root-mean square error of cross validation set（ RMSECV） was 0. 069 8. Besides,in the quantitative model of ferulic acid,the correlation coefficients（ R2） of calibration set and cross validation set were 0. 986 879 and 0. 962 243. RMSEC was 1. 402 6 and RMSECV was 2. 400 2.When the established models were applied to on-line monitoring,the correlation coefficients of predicted results and measured values for total solid content and ferulic acid were 0. 993 3 and 0. 991 6; root-mean square error of predicted value（ RMSEP） was 0. 039 3and 1. 669 3 respectively; mean relative deviation of predicted value（ RSEP） was 3. 49% and 3. 58%. The results indicated that the established models can be used to fast measure the contents of the total solid and ferulic acid in extraction process of Tianshu capsule extracts.