汽油替代物辛烷值正向建模及敏感性分析  

Forwarding Prediction Modeling and Sensitivity Analysis Study on the Octane Number of Gasoline Surrogates

作　　者：马寅杰 王勇[1] 刘成 鄂加强[1] MA Yin-Jie;WANG Yong;LIU Cheng;E Jia-Qiang(College of Mechanical and Vehicle Engineering,Hunan University,Changsha 410082,China)

机构地区：[1]湖南大学机械与运载工程学院,长沙410082

出　　处：《工程热物理学报》2021年第6期1608-1613,共6页Journal of Engineering Thermophysics

基　　金：国家自然科学基金资助项目(No.52006058);湖南省自然科学基金资助项目(No.2020JJ5071)。

摘　　要：基于辛烷值正向预测原理,本文发展了一种多项式混沌展开-全局敏感性分析(PCE-GSA)融合的高效燃烧参数分析方法,研究了初始温度、压力及汽油典型成分对辛烷值的影响规律。结果表明,采用PCE替代正向预测方案,可在保证预测精度的前提下,将建模需求的样本数量降低1~2个量级。在敏感性分析方面,初始热力学环境参数对辛烷值的Sobol敏感度随汽油种类、测试工况的不同而改变;相较于热力学环境,燃料成分的小范围扰动对辛烷值的影响不显著;乙醇的加入将明显地提升环境温度对汽油自燃特性的影响能力,同时进一步抑制了燃料成分对辛烷值的贡献度。Based on the principle of forwarding prediction,this paper proposed an efficient parameter analysis algorithm that combining polynomial chaos expansion and global sensitivity analysis method(PCE-GSA),and investigated the influence of initial temperature,pressure and the typical components in gasoline fuels on the octane number.The results show that adopting polynomial chaos expansion would substantially reduce the required sample size for modeling studies by 1 to 2 orders of magnitude in comparison with the conventional forward prediction method.In the aspect of sensitivity analysis,the Sobol sensitivity indices of initial thermodynamic parameters on fuel octane numbers change according to the gasoline types and test conditions;compared with the thermodynamic environment,the small-scale disturbances of fuel components would lead to an insignificant influence on the octane numbers.The addition of ethanol would enhance the influence of environmental temperature on gasoline’s octane number,and inhibit these contributions from the fuel components.

关 键 词：正向预测 辛烷值 多项式混沌展开 全局敏感性分析 

分 类 号：TK418.9[动力工程及工程热物理—动力机械及工程]