车用柴油关键性能指标的中红外光谱快速筛查  

Rapid screening of key performance indicators for automotive diesel using mid infrared spectroscopy

作　　者：苏玉倩 孙牧 马东 薛彦军[1] SU Yuqian;SUN Mu;MA Dong;XUE Yanjun(Jiangsu Provincial Institute of Product Quality Supervision and Inspection,Nanjing Jiangsu 210007,China)

机构地区：[1]江苏省产品质量监督检验研究院,江苏南京210007

出　　处：《石油化工应用》2023年第12期99-104,共6页Petrochemical Industry Application

摘　　要：选取具有代表性的、能够覆盖质量指标范围的车用柴油样品集合,即样品集,利用中红外燃油分析仪,以空气为参比,采集样品集背景光谱。利用化学计量学软件,以偏最小二乘法(PLS)建立各项质量指标与光谱数据关系的定标模型,用定标样品集的统计标准偏差(SEC)评价定标模型的准确性,以SEC是否满足参考标准方法的再现性进行评价。使用定标样品集外的样品验证定标模型的准确性,通过与标准检测方法进行对比实验,验证定标模型是否满足标准方法中规定的再现性要求,并进一步用统计学中配对t检验进行评价,验证定标模型是否与标准检测方法存在显著性差异。Select a representative set of automotive diesel samples that can cover the range of quality indicators,namely the sample set,and use a mid infrared fuel analyzer to collect the background spectrum of the sample set with air as the reference.Using chemometrics software,a calibration model for the relationship between various quality indicators and spectral data is established by partial least squares(PLS).The accuracy of the calibration model is evaluated by the statistical standard deviation(SEC)of the calibration sample set,and whether SEC meets the reproducibility of the reference standard method is evaluated.Verify the accuracy of the calibration model using samples outside the calibration sample set.Conduct comparative experiments with standard detection methods to verify whether the calibration model meets the reproducibility requirements specified in the standard method.Further evaluate using paired t tests in statistics to verify whether there is a significant difference between the calibration model and the standard detection method.

关 键 词：中红外光谱 定标模型 关键性能指标 准确性 快速筛查 

分 类 号：O657.3[理学—分析化学]