人工神经网络紫外光谱法测定氯芬待因片含量

目的:对紫外光谱重叠的氯芬待因片进行多组分不经分离的含量测定.方法:应用前向型误差反向传播人工神经网络对氯芬待因片进行含量测定.结果:3个模拟样品平均回收率和RSD分别为磷酸可待因98.68%,3.40%;双氯芬酸钠100.01%,0.94%.结论:该方法测定结果准确,性能良好,对于吸收光谱重叠的药物含量测定有一定的参考价值.

The Artifical Neural Networks Spectrophotometry Used in Simultaneous Determination of Two Ingredients in Shuerfen Tablet

Objective:To determine the concentration of the contents of shuerfen tablets which UV spectrum are overlapped. Method:The error back-propagation(BP) method of artifical neural networks(ANN) has been used to determine codeine phosphate and diclofenac sodium without separating them. In 400-190 nm, the absorbance at 9 wavelength was taken as character parameter of ANN,and samples value were input. Results:The average recoveries and relative standard deviations of two components in three model samples were 98.68%, 3.40% for codeine phosphate and 100.01%, 0.94% for diclofenac sodium respectively. Conclusion:The method is accurate and the network performs satisfactorily, and the method could be used as reference for determining drugs which UV spectrum are overlapped.