【摘要】 本文讨论了我国润滑油基础油粘度指数与红外光谱信息的关系,应用计算机模式识别技术进行分类处理,建立了我国润滑油基础油粘度指数分类特征集,并通过了分类器的检验。研究结果表明,基础油的粘度指数与其红外光谱信息之间有密切关系;在模式空间中粘度指数的变化趋势明显;基础油的粘度指数分布有相当明显的地区性。本文还发展了一种基于k-NN法的半定量粘度指数估算方法,用以解决一般模式识别技术难以解决的估值问题。更多还原

【Abstract】 The relationship between IR information and the viscosity index (Ⅵ) of Chinese lubricating base oils was discussed in this article with Pattern Recognition Techniques. The distribution maps of Ⅵ of samples were given. It is found that these distribution maps of Ⅵ are strongly related to their oil sources. A suitable method based on the k-NN classifier to estimate Ⅵ of unknown patterns is also developed.更多还原