基于峰值拟合算法的光谱分析

Spectra Analysis Based on Peak Fitting Algorithm

傅里叶变换光谱学已经用于环境气体的定量分析中,为了提供高灵敏、连续实时的在线测量,分析中需要根据实测光谱选择光谱拟合算法。文中将峰值拟合算法应用于气体红外光谱分析中,以实际测量得到的气体透过率光谱为拟合对象,将其与数据库中的气体分子模型光谱进行拟合,在算法中通过迭代使残差的加权平方和最小,获得了气体浓度等重要的光谱拟合结果。算法实现结果表明该方法不仅精确可靠而且具有简单快速的特点。

Fourier transform infrared spectroscopy has been employed in environmental gas quantitative analysis. In order to provide high sensitivity and continuous real - time on - line measurements, some key techniques such as fitting algorithms should be choosed to analyze the measured spectra. Peak fitting algorithm is applied to gaseous infrared spectra analysis in the paper. The measured gas transmittance spectra are taken as fitted object. They are fitted with model spectra in molecular database. The weighted sum square of residuum is required to be the least one in the iteration of algorithm. The important results such as gas concentration are acquired by fitting. The implementation of the algorithm has shown that the method is not only exact and reliable but also simple and rapid.