当前位置：首页 > 过刊浏览->2026年第37卷第4期
研究法辛烷值近红外模型在汽油生产中的应用
Application of Near-Infrared Spectroscopy Model for Research Octane Number in Gasoline Production
浏览(0) 下载(0)
- DOI:
- 作者:
- 李建国
Li Jianguo
- 作者单位:
- 中韩石化检验中心
Inspection Center of China-South Korea Petrochemical
- 关键词:
- 近红外光谱法；建模；汽油调和；研究法辛烷值
near-infrared spectroscopy; model calibration; gasoline blending; Research Octane Number
- 摘要:
- 以近红外法预测汽油研究法辛烷值（RON）为例，介绍近红外建模和应用全过程。为优化汽油生产过程，建立近红外模型预测汽油研究法辛烷值，运用该模型预测40个模型以外汽油样品的研究法辛烷值，并与方法GB/T 5487-2015测量值进行对比，模型预测总体正确率达97.5%，该方法快速、准确，满足汽油生产调和的需求。通过建立近红外模型提高汽油一次调和合格率，有效控制研究法辛烷值数据，避免了汽油的质量过剩，同时提高了分析效率，节省化验分析成本。Taking the prediction of gasoline Research Octane Number (RON) by near-infrared spectroscopy as an example, this pa⁃per introduces the whole process of near-infrared model establishment and application. To optimize the gasoline production process,a near-infrared model was built to predict the research octane number of gasoline. The model was adopted to predict the RON of 40gasoline samples outside the modeling set, and the predicted values were compared with the measured results obtained in accor⁃dance with GB/T 5487-2015. The overall prediction accuracy of the model reaches 97.5%. Being rapid and accurate, this methodcan meet the requirements of gasoline blending in production. The established near-infrared model helps improve the one-passqualification rate of gasoline blending, realize effective control of research octane number, avoid over-specification of gasoline quali⁃ty, increase analytical efficiency and reduce laboratory testing costs.