论文部分内容阅读
盲信号分离算法可以实现重叠化学信号分离。在色谱分析中对盲信号分离命题进行解释,提出独立成分分析的独立成分对应分离的单位色谱峰,混合矩阵的元素对应各组分含量的单位比值。构造两组分的模拟色谱重叠峰作为计算验证,测定正已烷/正庚烷混合体系的色谱重叠峰作为实验验证。经过模拟和实验验证,对色谱分析中的重叠峰,进行各组分比例不同的多次测定,测定次数大于或等于组分数目,采用信号盲分离算法可以求出各组分的单位色谱峰,并确定重叠峰内各组分的相对含量。
Blind signal separation algorithm can achieve overlapping chemical signal separation. In the chromatographic analysis, we explain the proposition of blind signal separation, and propose that the independent components of the independent component analysis correspond to the unit chromatographic peaks separated, and the elements of the mixed matrix correspond to the unit ratios of the content of each component. As a computational verification, the overlapping peaks of the two chromatographic structures were constructed and the chromatographic overlapping peaks of n-hexane / n-heptane mixtures were determined as experimental results. After simulation and experimental verification, the overlapping peaks in the chromatographic analysis were tested with different proportions of different components, and the number of determination times was greater than or equal to the number of components. The unit chromatographic peaks of each component were obtained by signal blind separation algorithm, And determine the relative content of each component within the overlapping peaks.