首页|喷气燃料烃类组成与性质构效关系研究进展

喷气燃料烃类组成与性质构效关系研究进展

扫码查看
通过喷气燃料分子水平研究和表征,掌握喷气燃料烃类组成,进而与理化性质进行构效联系,对喷气燃料性质预测、快评技术研发和高性能喷气燃料设计等具有重要指导作用.综述了喷气燃料的燃烧性、挥发性、低温流动性、安定性、润滑性、导电性、洁净性和燃料烃类组成与常规理化性质的定性关联规律,归纳了具有代表性的定量关系模型.总结了燃料的化学组成分析手段,认为全二维气相色谱应用于喷气燃料烃族及碳数分布等分子水平表征更准确.展望了预测模型的开发方向,指出相比于现有多元线性回归(MLR)、加权平均(WA)、偏最小二乘(PLS)、修正加权平均(MWA)、支持向量机(SVM)等建模方法,人工神经网络(ANN)是未来定量关系模型的研究重点.
Research Progress on Structure-Activity Relationship Between Hydrocarbon Compositions and Physicochemical Properties of Jet Fuel
Through the research and characterization of jet fuel molecular level,the hydrocarbon compositions of jet fuel are determined to further build the structure-activity relationship with physical and chemical properties,which will play an important role in guiding the prediction of jet fuel properties,R&D of rapid evaluation technology and design of high-performance jet fuel.The combustion,volatility,low-temperature fluidity,stability,lubricity,conductivity and cleanliness of jet fuel are comprehensively reviewed,as well as the qualitative correlation law between hydrocarbon compositions and general physicochemical properties,and the representative quantitative mathematical models are summarized.Meanwhile,based on summarizing the analysis methods of fuel chemical compositions,it is considered that the comprehensive two-dimensional gas chromatography is more accurate in molecular level characterization of the hydrocarbon group and carbon number distribution of jet fuel.Moreover,the development direction of prediction model is prospected and the results indicate that compared with the current modeling methods such as multiple linear regression(MLR),weighted average(WA),partial least square(PLS),modified weighted average(MWA),support vector machine(SVM),and artificial neural network(ANN)are a research focus for the future.

jet fuelhydrocarbon compositionphysicochemical propertymath model

蔡璐、舒亦桥、陶志平、赵杰、伏朝林

展开 >

中石化石油化工科学研究院有限公司,北京 100083

喷气燃料 烃类组成 理化性质 数学模型

中国石油化工股份有限公司课题

ST20036-6

2024

石油学报(石油加工)
中国石油学会

石油学报(石油加工)

CSTPCD北大核心
影响因子:0.764
ISSN:1001-8719
年,卷(期):2024.40(1)
  • 20