结合非对称最小二乘法与傅立叶变换对差分吸收光谱的解析

Analysis of Differential Absorption Spectrum by Integrating Asymmetric Least Squares Method with Fourier Transformation

在分析差分吸收光谱技术中的光谱时,光谱慢变化部分拟合结果的好坏尤其关键。在光谱处理过程中,提出了结合非对称最小二乘法与傅立叶变换来解析气体吸收光谱的方法。气体的吸收光谱经傅立叶变换后得到其对应的频谱曲线,在频谱的基础上,借助非对称最小二乘法拟合出频谱中对应于吸收光谱中慢变化部分的基线。根据傅立叶变换的性质,在频谱的函数值上减去所拟合的基线值,达到去除吸收光谱中的慢变化部分的目的;进一步处理得到气体的差分吸收光谱,从而反演出气体浓度;并与传统的光谱处理方法——多项式拟合光谱慢变化的处理结果做了比较,验证了结合非对称最小二乘法与傅立叶变换来解析差分吸收光谱分析方法的可行性和优越性。

The result of the slow varying fitting is pivotal in analyzing the differential absorption spectrum. A method based on the integration of asymmetric least squares method （ALSM） with Fourier transformation （FT） was developed for the analysis of the differential optical absorption spectrum （DOAS） . The absorption spectrum of the gas was first Fourier transformed into the frequency spectrum . Based on the frequency spectrum and the ALSM, the slow varying part of the baseline of the absorption spectrum was fitted. The fitted baseline was then subtracted from the function values of the frequency spectrum to remove the slow varying part of the absorption spectrum. The gas concentration was finally inversed by further processing the DOAS of the gas. Compared with the traditional polynomial fitting of the slow varying part of the spectrum, the proposed method was beneficial and feasible.