一种新型的信号拟合方法─—免疫算法

A Novel Algorithm for Fitting Analytical Signals──an Immune Algorithm

通过对免疫系统免疫机制的模拟，提出了一种新型的信号拟合算法。该算法根据所提供的标样信息对重叠分析化学信号进行拟合，从重叠信号中提取单一组分的信息，从而实现了多组分混合信号的解析。对模拟信号和实验信号的处理结果表明，该方法可方便地用于多组分重叠色谱信号的解析。

A novel immune algorithm for fitting analytical signals was proposed by imitating the mechanism of the immune system, and application of the algorithm to the resolution of overlapping chromatograms was investigated. The method takes an overlapping chromtogram as its input and iteratively subtract the estimated chromatographic information of every cpmponent from the input until the residual does not change. Both simulated and experimental data sets were investigated by the method. It was proved that the method is a convenient tool for resolution of overlapping chromatograms.