摘要
介绍了汽油组分及汽油辛烷值预测方法,重点阐述拓扑指数法、基团贡献法对汽油组分辛烷值的定量结构-性质相关性的研究,并综述利用仪器分析提取参数对汽油辛烷值进行构效关系研究。综述了分别采用人工神经网络法﹑遗传算法﹑支持向量机﹑多元线性回归法﹑偏最小二乘法对汽油及汽油组分辛烷值进行拟合的情况及优劣,由于拓扑指数计算比较简单,能较完善的表示分子结构特征,得到的辛烷值预测模型的相关性比其他方法较好,该方法是目前比较广泛使用的一种方法。
The prediction methods of octane number lor gasohne components and gasohne were mtrouuceti, lne tocus was on using the topological index method and group contribution method to study the quantitative structure-property relationship （QSPR） of octane number of gasohne components and using the parameters extracted by instrument to study QSPR of octane number of gasoline. The advantages and disadvantages of various methods employed for fitting, including artificial neural network, genetic algorithm, support vector machine, multiple linear regression and partial least squares, were analyzed. Topological indices had relatively simple calculation process and could more completely represent the molecular structures, and the relevance of the prediction model produced by it was better than those by other methods. Therefore, topological index method was most widely used.
出处
《天然气化工—C1化学与化工》
CAS
CSCD
北大核心
2014年第2期62-66,共5页
Natural Gas Chemical Industry
基金
广东省部产学研结合项目(?2009B090300134)
广东省高新区引导专项项目(2011B010700060)
2010年广东省重大科技专项子项目(911009)
关键词
辛烷值
汽油
汽油组分
拓扑指数法
基团贡献法
octane number
gasoline
gasoline component
topological index method
group contribution method
