【摘要】 本文利用判别法模式识别技术,对我国一些润滑油基础油(包括部分国外基础油)的三种氧化安定性模拟评定结果与基础油的族组成,结构族组成以及总硫、总氮和碱性氮含量之间的关系进行了探讨,建立了较为合适的特征集,分类效果明显。本文工作表明,计算机模式识别技术是辅助石油产品研制的有力工具。更多还原

【Abstract】 Discriminant Pattern Recognition Techniques were adopted to study the relationship between oxidation stability and composition of domestic lube base oils (some foreign base oils were also examined).Oxidation stabilities of base oils were determined separately by three different laboratory simulated testing methods (ASTM D-2272, SY 2681, and dynamic oxygen absorption). The relevant composition information of base oils includes group composition, structural group composition and the content of total sulfur, total nitrogen and basic nitrogen.Appropriate feature sets were constructed.The results of classifying patterns in training sets and recognizing of unknown patterns are evident.It shows that Pattern Recognition Techniques can be a powerful tool to assist the development of petroleum products.更多还原