人工神经网络-近红外光谱法非破坏定量分析Norvasc药片

Nondestructive Quantitative Analysis of Norvasc in Tablets by Artificial Neural Network Method-near Infrared Spectroscopy

用人工神经网络(ANN)解析Norvasc药片的近红外(NIR)漫反射光谱,实现了对其中有效成分苯磺酸氨氯地平精确的非破坏定量测定.设计了最佳的ANN模型,分别讨论了输入层和隐含层节点数对测定结果的影响.以HPLC法的测定结果为标准,苯磺酸氨氯地平浓度预测值的相对误差RE(%)<4.2%.

Artificial spectrum available of Norvasc in Norvasc ANN was designed. neural network （ANN） algorithm was used to analyse the near infrared （NIR） reflectance in tablet medicine. The content of amlodipine besylate, which was pharmaceutically medicine, was accurately nondestructively quantitatively determined. The best model of The effects of input nodes and hidden nodes on the results were discussed, respectively.Compared the results with those of HPLC, the relative error （RE%） of mlodipine besylate was less than 4.2%.