基于干涉图解析的被动式遥感FTIR分析

Interpretation of Passive Remote Sensing FTIR Spectra Based on Interferogram

针对被动式遥感傅里叶变换红外光谱(FTIR)在实际应用中存在受环境干扰较大,检测信号较弱的问题,利用背景和样品干涉图的衰减速度不同,建立了基于干涉图对被动式遥感FTIR谱图进行分析的方法。选择有限脉冲响应(FIR)滤波器提取信号,带宽选择为20 cm-1,干涉图长度为100点,距离中心爆发点(centerburst)第50点为干涉图起始点,对样品苯的被动式遥感FTIR信号进行了有效提取。然后采用偏最小二乘法(PLS)建立模式识别模型,对26个未知样品的干涉图数据进行预测,总识别率为96%。该方法的建立,减弱了背景对遥感测试的影响,强化了被动式FTIR的分析信号,同时与化学计量学方法结合实现了被动式遥感FTIR对污染物的自动检测。

In view of background interference and weak signals in practice,a method for analyzing passive remote sensing FTIR signals with interferogram was built based on different attenuation rates of background and sample interferogram.With a bandwidth of FIR filter of 20 cm^-1,the signal of benzene was acquired from the passive remote sensing FTIR interferogram while parameters were optimized as follows：interferogram length was 100 points,and the start point of interferogram was 50 points apart from the centerburst.Then PLS was utilized to build a calibration model and recognize 26 unknown samples with a recognition rate of 96%.The work indicates that the responses of passive remote sensing FTIR were intensified and the background effect was avoided with this technique.Combination of this technique with chemometric method makes if possible to monitor the environmental pollution automatically.