论文部分内容阅读
本文应用偏最小二乘法(PLS)同短波近红外漫反射光谱相结合,对复方头孢氨卞片进行了无损非破坏定量分析,以样品中头孢氨卞和甲氧卞啶为活性成分建立了最佳的数学校正模型。分别讨论了三种预处理光谱,即标准归一化(SNV)、一阶导数和二阶导数光谱中主成分数对PLS定量预测能力的影响,证明使用SNV预处理光谱更有优越性。
In this paper, partial least square method (PLS) combined with shortwave near-infrared diffuse reflectance spectroscopy was used to perform nondestructive and non-destructive quantitative analysis of compound cefmenoxime tablets. The optimal composition of cefmenoxime and trimethoprim Mathematical correction model. The effects of three kinds of pretreatment spectra, namely, the normalized normalization (SNV), the first derivative and the second derivative spectra on the quantitative prediction ability of PLS, respectively, are discussed. It is proved that SNV pretreatment spectrum is more superior.