多元回归在润滑油生产中的应用  

Application of Multiple Regression in Lubricant Production

作　　者：何国成[1] 李会云[1] HE Guo-cheng;LI Hui-yun(Northwest Branch of Sinopec Lubricant Company,Zhengzhou 450001,China)

机构地区：[1]中国石化润滑油西北分公司,河南郑州450001

出　　处：《润滑油》2021年第6期60-64,共5页Lubricating Oil

摘　　要：依据混合黏度计算模型,根据灌装油品、管线残留油品100℃运动黏度,利用多元回归确定各段管线末次残留油品对灌装首检结果的影响,并通过相关系数的显著性检验、F检验和t检验;通过在调合结束后增加回流的工艺改进,减少顶线量,减少顶线油使用的质量风险。The effect of last residual oil in each pipeline section on the first inspection results of filling was investigated through multiple regression.The viscosity calculation model based on the kinematic viscosity at 100℃of filling oil and residual oil in pipeline was employed,and the significance test,F test and t test of correlation coefficient were passed.An improved technique,which increases backflow after blending,was employed to reduce top line volume and the quality risk of using top line oil was decreased.

关 键 词：黏度 回归 线性 工艺 顶线 

分 类 号：TE626.3[石油与天然气工程—油气加工工程]