支持向量回归-紫外分光光度法用于测量小儿氨酚匹林咖啡因片含量的方法研究

目的建立同时测定小儿氨酚匹林咖啡因片中3种组分:阿司匹林、对乙酰氨基酚和咖啡因含量的紫外光谱的支持向量回归校正方法.方法对复方制剂的紫外光谱数据进行预处理和主成分分析后不经分离,采用支持向量回归(SVR)算法同时测定3组分的含量.结果测定正交设计标样中的阿司匹林、对乙酰氨基酚和咖啡因的平均回收率分别在98.2%～101.2%之间,RSD在0.47%～0.91%之间.将SVR法与偏最小二乘回归和径向基神经网络建模方法相比较,SVR所建模型的预测准确性优于后两者.结论本法可用于小儿氨酚匹林咖啡因片的紫外光谱的含量测定分析.

UV spectrometry and support vector regression for simultaneous determination of paracetamol,aspirin and caffeine

Objective To use multivariate spectrophotometric calibration for the simultaneous analysis of paracetamol,aspirin and caffeine in tablets for children.Methods After UV data were pretreated and the principal component analyzed,support vector regression(SVR) was employed to analysis the three components with a high degree of spectral overlap.Results The proposed method was compared with partial least square regression and radial-basis function neural network modeling methods.The predictive accuracy of UV calibration models built by SVR was much better than that of the models built by partial least square regression and radial-basis function neural network.The average recoveries for the three components were between 98.2% and 101.2%(n=49) and RSD were between 0.47% and 0.91%(n=49).Conclusion The method could be used for simultaneous analysis of the contents of paracetamol,aspirin and caffeine in tablets for children.