遗传算法用于傅里叶变换红外光谱的定量解析

本文利用遗传算法对混叠的傅里叶变换红外光谱图进行了定量的解析。文中分析了最多包含了十个组分的大气有毒有机物（苯，四氯化碳，氯苯，硝基苯，苯酚，甲苯，甲醇，邻甲酚，间甲酚和对甲酚）的混合谱图，建立了适当目标函数，使用了自适应变异技术，讨论了交叉和变异概率对结果的影响。结果证明遗传算法有着极好的非线性特征。在适当的条件下，结果的相对误差小于1％。

Using Genetic Algorithm for Quantitative Analysis of Overlapped Spectra in FTIR

Technique for the quantitative analysis of FTIR spectra,which are serious overlapped in the spectral bands,is studied by genetic algorithm.The multicomponent in the mixture,which contains ten air toxic organic components,such as benzene,phenixin,chlorobenzene,nitrobenzene,phenol,toluene,methanol,o-cresol,m-cresol and p-cresol,is simultaneously determined.The effect of the probability of crossover and the establishment of objective function are discussed.The method of adaptive mutation is used.The results indicate that gentic algorithm has excellent non-linear ability of solution.Mean relative error of each results is really small,generally not bigger than 1%,when the conditions are properly controlled.