【摘要】 提出了用高维PCA从未知样及校准样组合的三维阵中提取最具代表性抽象光谱及浓度信息的想法.并结合QZ算法提出了一种新的减秩法(NRA)。实际用于混合物荧光定性定量分析中,发现它比广义减秩法(GRA)结果稳定、可靠。光谱重叠、噪声等环境因素的干扰大大减小。更多还原

【Abstract】 A new idea of extracting most representative abstract spectra and concentration information from a three dimensional matrix composed of an unknown EEM and a calibration EEM through muiti-dimensional PCA was presented in this paper. By combining these informations with QZ algorithm, a new rank annihilation method (NRA) was proposed. It has been used practically in qualitative and quantitative analysis of fluorescent mixtures and a comparison was made with generalized rank annihilation method (GRA). It was found that the results of NRA were more stable and reliable. The influence of environmental factors such as spectra overlap,noise,etc. was sharply reduced.更多还原