递推偏最小二乘法的油品芳烃烯烃成分建模

Recursive Partial Least-squares Modeling of Aromatics and Olefins in Gasoline

采用递推偏最小二乘法建立了适合测定油品中芳烃和烯烃含量的分析模型,并在某炼油企业应用达到了很好效果;近红外光谱法的测定结果与荧光指示剂吸附法的测定结果具有很好的一致性,与荧光指示剂吸附法相比,近红外光谱法可以提高分析效率,降低分析成本,具有较高的分析精密度。

The separate calibration models of aromatics and olefins were established for gasoline through recursive partial least square （R - PLS） method, the application has achieved better effect in some oil refining enterprises. The calibration models were validated through comparison of the results determined by fluorescent indicator adsorption （FIA） and near infrared spectroscopy （NIR） methods. The NIR analysis results were well coincident with those of FIA method. The NIR can not only raise the analysis efficiency and lower the analysis cost,but also has better precision compared with FIA method.