【作者】 虞正亮；

【导师】 邵学广；

【作者基本信息】 中国科学技术大学 ， 分析化学， 2006， 博士

【摘要】 多组分重叠信号的解析一直是分析化学研究领域中具有挑战性的问题，免疫算法和小波变换是两种解析多组分重叠信号的有效工具。本文首先简要介绍了免疫算法和小波变换的基本原理和基本算法，并综述了近年来在分析化学领域中的应用。然后对免疫算法和小波变换在解析重叠信号方面的应用进行了深入研究，完成了以下几个方面的工作内容：基于小波压缩和免疫算法，提出了一种用于解析多组分重叠二维色谱信号的快速算法。通过对多组分二维重叠色谱信号的解析和定量计算，结果表明小波压缩—免疫算法的计算速度快、定量准确，是解析多组分二维重叠色谱的有效方法，可作为免疫—遗传算法的快速预处理的候选方法。 基于免疫算法的基本原理，提出了一种自适应免疫算法。通过建立标准峰的数据库，从数据库中自适应地产生用于混合物信号解析的标样信息。通过计算混合物信号与标样信号之间的相关系数和采用优化策略，使算法变得更加智能，算法的效率得到明显提高。将其应用于4组分NMR谱和9组分NMR谱的解析，结果表明该方法是一种有效的自适应免疫算法，与免疫遗传算法相比计算速度更快，解析结果更精确。 为了对多组分重叠色谱信号进行定量分析，提出了将变形算法与小波变换平滑、三次样条插值及一个新的确定最佳数据点数的标准相结合的改进算法。使用改进的算法，可以有效抑制由于变形产生的振荡。通过对模拟信号的解析，结果表明改进的算法具有更强的解析能力，降低了对重叠信号分离度、噪声水平及峰形方面的限制要求。通过对实验色谱信号的解析，进一步证明了其优点。研究结果表明，改进的变形算法与原来的算法相比在多组分重叠色谱定量解析方面更具有实用性。 本文还对独立成分分析法(ICA)在多组分重叠NMR解析中的应用进行了研究。由于NMR信号与ICA假设条件之间的不一致，在某些情况下使用ICA不能直接得到正确的解析结果。因此，提出了一种对ICA的计算结果进行再旋转的方法用来修正ICA的计算结果。通过对模拟10组分混合物NMR谱和2组分更多还原

【Abstract】 Resolution of multicomponent overlapping signals is one of the challenging tasks in analytical chemistry studies. Both immune algorithm and wavelet transform are effective tools for resolution of multicomponent overlapping signals. In this dissertation, at first, the theory and algorithm of immune algorithm and wavelet transform are introduced and the applications of them in analytical chemistry are reviewed. Then, with the aim to develop new tools for resolution of overlapping analytical signals, the following works are carried out:Based on the wavelet compression and immune algorithm, a fast algorithm for resolution of 2-D multicomponent overlapping chromatogram is proposed. By application of the method in resolution and quantitative determination of multicomponent 2-D overlapping chromatograms, it is demonstrated that this method is fast in calculation speed and accurate in quantitative calculation. Therefore, the proposed algorithm may be an alternative effective method for resolution of multicomponent 2-D overlapping chromatogram.Based on the immune algorithm, an adaptive immune algorithm is proposed. With the establishment of the database of the standards peaks, the standard information is adaptively generated from the database for the resolution of the mixture sample. With the adoption of the correlation coefficient between mixture signal and standard signals and the optimal strategy, the algorithm is much more intelligentized and the efficiency of the algorithm is improved obviously. From the resolution results of a four-component NMR spectrum and a nine-component NMR spectrum, it is demonstrated that the proposed algorithm is an effective adaptive immune algorithm. Compare with the IA-GA, the calculation speed of using the AIA is much faster and the resolved result is much more accurate.With the aim of quantitative determination of multicomponent overlapping chromatograms, an improvement of the transmutation method was proposed with WT smoothing, cubic spline interpolation, and a new criterion for determination of the optimal points number. With the proposed algorithm, the oscillation generated in the transmutation process can be effectively suppressed. By the application to simulated更多还原