论文部分内容阅读
根据待测物随pH变化的分布分数曲线并结合其吸收光谱构成两维波谱数据矩阵,利用约束背景双线性分解法(CBBL)可从未知干扰背景的灰色体系中对其进行定量分析。通过模拟实验,对待测物各型体间的波谱重叠程度及待测物与干扰物的分布分数曲线之间的波谱重叠程度对测定准确度的影响进行了研究。确定了可对待测物准确地进行定量分析时分析体系中波谱间的波谱重叠程度值。对酪氨酸、色氨酸和苯丙氨酸三种氨基酸的混合样的测定结果验证了预报准确度与波谱重叠程度之间的依赖关系。
The two-dimensional spectral data matrix is constructed according to the distribution curve of the analyte with the change of pH and its absorption spectrum. The constrained background bilinear decomposition (CBBL) can be used to quantitatively analyze the data from the gray system with unknown interference background. Through simulation experiments, the influence of the degree of spectral overlap between the various types of the analyte and the degree of spectral overlap between the distribution curves of the analyte and the interfering substance on the determination accuracy was studied. The value of spectral overlap between spectra in the analytical system can be determined when the analyte is accurately quantified. The results of a mixed sample of three amino acids, tyrosine, tryptophan and phenylalanine, confirm the dependence of prediction accuracy on the degree of spectral overlap.