多种分子光谱快速分析黄芩中的黄芩苷含量

发布时间：2018-03-14 11:10

本文选题：多源复合光谱仪　切入点：近红外光谱　出处：《分析测试学报》2017年05期 　论文类型：期刊论文

【摘要】：为考察根据不同类型光谱信息进行黄芩质量快速分析的适应性,采用高效液相色谱(HPLC)法测定了73批黄芩样品中的黄芩苷含量并作为y值,以各样品的近红外、紫外-可见光谱及包含紫外、可见及近红外的多源复合光谱信息作为x值;根据各类光谱信息分别采用偏最小二乘回归(PLSR)与K最近邻样本保形映射(KNN-KSR)方法进行样品中黄芩苷的预测,根据验证集样本真实值与预测值的均方根偏差(RMSEP)、平均相对误差(MRE)与相关系数(R)评价预测精度。结果表明,采用KNN-KSR方法根据各类光谱信息预测黄芩苷时,各项指标均优于PLSR方法的结果;其中基于近红外光谱对黄芩苷的分析结果最好,紫外-可见光谱次之,基于多源复合光谱信息对黄芩苷的预测误差最大,但其MRE仍在6%以下,可满足工业分析的精度要求。由于多源复合光谱仪具有体积小、重量轻、成本低及便携等优点,通过优化仪器波长范围及建模方法,有望改进该仪器的分析精度,使之适应更多药材现场采购的快速检测及后续产品的质量分析与监控需求。
[Abstract]:In order to investigate the adaptability of rapid quality analysis of Scutellaria baicalensis based on different spectral information, the content of baicalin in 73 batches of Scutellaria baicalensis samples was determined by high performance liquid chromatography (HPLC). The UV-Vis spectrum and the multi-source composite spectral information containing UV, Vis and NIR are taken as x values. According to various spectral information, partial least square regression (PLSRs) and K nearest neighbor sample conformal mapping (KNN-KSRs) were used to predict baicalin in samples. The prediction accuracy was evaluated according to the root mean square deviation (RMS EPN), mean relative error (Mre) and correlation coefficient (R) of the true and predicted values. The results show that the KNN-KSR method is used to predict baicalin based on various spectral information. The analysis results of baicalin based on near infrared spectroscopy were the best, the UV-Vis spectrum was the second, and the prediction error of baicalin based on multi-source composite spectrum information was the biggest, but the MRE of baicalin was still below 6%. The multisource composite spectrometer has the advantages of small volume, light weight, low cost and portable. By optimizing the wavelength range of the instrument and modeling method, it is expected to improve the analytical accuracy of the instrument. To adapt to more rapid detection of field procurement of medicinal materials and the quality analysis and monitoring requirements of subsequent products.
【作者单位】：华东理工大学化学与分子工程学院;
【基金】：上海市科学技术委员会支撑项目(13401901100)
【分类号】：O657.3;R284.1

【相似文献】