论文部分内容阅读
中药材提取过程是感冒灵颗粒剂生产过程的首个环节,对药品质量影响显著。该文在线采集感冒灵中药材提取过程近红外光谱图,以高效液相色谱和差重法为参照方法,采用一阶导数法处理近红外光谱,运用偏最小二乘回归法(PLSR)分别建立提取液中蒙花苷、绿原酸和固含量的定量校正模型。采用相关系数(r)、交叉验证均方差(RMSECV)、校正集均方差(RMSEC)和验证集均方差(RMSEP)等指标优化建模参数,考察模型性能。3种质量控制指标的模型相关系数均达到0.95以上,蒙花苷和绿原酸、固含量的RMSEC和RMSEP分别为0.010 4和0.009 47,0.009 34和0.142,0.055 5和0.008 42,在线分析所建模型,预测值与实际测定值相关系数均大于0.97,其预测相对偏差(RSEP)分别为8.14%,8.17%,9.86%。研究结果表明,利用近红外光谱技术可以实现感冒灵中药提取过程多指标的在线检测和实时监控,该技术可用于生产过程中质量控制,缩小中间体批次差异性,保证药品质量稳定性,也为后续的产品质量回溯提供了实时生产数据。
The extraction process of Chinese herbal medicines is the first step in the production process of Ganmaoling granules and has a significant impact on the quality of medicines. In this paper, the near infrared spectra of Chinese herbal medicines for the extraction of Ganmaoling were collected online. The first-order derivative method was used to process the near-infrared spectra with reference to the method of high performance liquid chromatography and differential multiple. The partial least squares regression (PLSR) Quantitative Calibration Model of Monatin, Chlorogenic Acid and Solid Content in Extracts. The model parameters were optimized by using correlation coefficient (r), RMSECV, RMSEC and RMSEP. The correlation coefficients of the three quality control indexes were above 0.95. The RMSEC and RMSEP of the solid content of Monurin and chlorogenic acid were 0.010 4 and 0.009 47,0.009 34 and 0.142,0.055 5 and 0.008 42 respectively. On-line analysis The correlation coefficient between the model and the predicted value and the actual measured value was more than 0.97, and the relative prediction error (RSEP) was 8.14%, 8.17% and 9.86% respectively. The results show that the near-infrared spectroscopy can be used to achieve online detection and real-time monitoring of multi-indexes in the process of Chinese herbal medicine for the detection of Ganmaoling. This technique can be used for quality control in the production process, narrowing the difference between batches of intermediates, ensuring the stability of drug quality, For the follow-up of product quality back to provide real-time production data.