近红外光谱快速测定石脑油芳烃含量模型的校正研究

Research on Calibration Model for Rapidly Determining Aroma Composition in Naphtha by Near-infrared Spectroscopy

在采用近红外光谱建立石脑油芳烃含量校正模型的过程中,分析讨论了校正系数K值、光谱区域选取及光谱预处理方法对校正模型准确性的影响。实际分析结果表明,选择11000cm^(-1)～6099．7cm^(-1)光谱区域并使用偏最小二乘法校正方法建立的石脑油芳烃含量模型的预测值偏差不大于2．0％,能够实现对石脑油芳烃含量的快速准确测定。该方法在实际生产过程中已得到应用。

The influence of calibration constant K, spectrum region selection and spectrum preprocessing on the veracity of a calibration model when establishing a model for rapidly determining aroma composition in naphtha by near-infrared spectroscopy are analyzed. The analysis result has shown that the model for rapidly determining the aroma composition in naphtha established by using least square in the spectral region from 11000cm^-1 to 6099.7cm^-1 has a predicted value less than 2.0% and can be used to determine the aroma composition in naphtha rapidly and accurately. Now, this method has been used in the production process.