论文部分内容阅读
目的:建立同时快速鉴别3种药典收载麻黄药材的近红外光谱法。方法:测量药典品种麻黄药材草麻黄、中麻黄和木贼麻黄,以及混淆品种丽江麻黄的傅里叶变换近红外漫反射光谱,采用化学计量学技术,提取不同药典品种麻黄药材共有的且区别于混淆品种丽江麻黄的光谱特征信息,建立同时鉴别不同药典品种麻黄药材与混淆品种丽江麻黄的判别分析(discriminant analysis,DA)模型和对向传播人工神经网络(counterpropagation artificial neural network,CP-ANN)模型,并比较2种模型的性能。结果:所建DA模型校正集和验证集的预测准确率均为100.0%;CP-ANN模型校正集、交叉验证和验证集的预测准确率均为100.0%。虽然2种模型具有相同的预测准确率,但是非线性CP-ANN模型使用较少的主成分建模,性能略优于线性DA模型。结论:所建近红外光谱法能够同时快速鉴别3种药典收载麻黄药材。
Objective: To establish a rapid and rapid identification of three pharmacopoeias containing ephedra ephedra by near infrared spectroscopy. Methods: The Pharmacopoeia ephedra ephedra herb ephedra, ephedra and aristichthys ephedra, as well as the confused variety Lijiang ephedra Fourier transform near-infrared diffuse reflectance spectroscopy, the use of chemometrics techniques to extract different pharmacopoeia Ephedra common and distinguished from The discriminant analysis (DA) model and the counterpropagation artificial neural network (CP-ANN) model were established to identify the spectral characteristic information of different varieties of Ephedra sinica and different varieties of Ephedra sinica simultaneously. , And compare the performance of the two models. Results: The predicted accuracy of the calibration set and validation set were both 100.0%. The prediction accuracy of the calibration set, cross-validation and validation set of CP-ANN model was 100.0%. Although the two models have the same prediction accuracy, the nonlinear CP-ANN model uses less principal component modeling and performs slightly better than the linear DA model. Conclusion: The proposed near-infrared spectroscopy can simultaneously identify three kinds of Pharmacopoeia containing ephedra medicinal materials.