加氢基础油抗氧化性能的模拟预测研究  

SIMULATED PREDICTION OF ANTIOXIDANT PERFORMANCE OF HYDROTREATING BASE OIL

作　　者：谢欣[1] Xie Xin(SINOPEC Research Institute of Petroleum Processing,Beijing 100083)

机构地区：[1]中国石化石油化工科学研究院

出　　处：《石油炼制与化工》2020年第2期57-61,共5页Petroleum Processing and Petrochemicals

基　　金：中国石油化工股份有限公司合同项目(107010)

摘　　要：针对不同加氢精制Ⅱ/Ⅲ类润滑油基础油(简称基础油)的结构特征,分别应用多层感知神经网络和径向基神经网络方法将加氢基础油链烷烃、环烷烃、烷基苯含量和黏度指数作为输入变量,建立了以氧化安定性为输出变量的9参数神经网络模型。首次将黏度指数作为预测模型输入参数,使模型预测准确度得到大幅提高。通过对影响加氢基础油氧化安定性的因素分析,找出了与氧化安定性有正/负相关关系的Ⅱ/Ⅲ类基础油的烷烃组成成分。在Ⅱ/Ⅲ类基础油中,在芳烃含量很低的情况下,随着链烷烃含量的增加,二环、三环环烷烃含量减少,润滑油的氧化安定性增加。According to the structural characteristics of hydrofiningⅡ/Ⅲbase oils from different producing areas,multi-layer perceptual neural network and radial basis function neural network were used to take the contents of alkanes,naphthenes,alkylbenzene and viscosity index of hydrorefining base oils as input variables,and a nine-parameter neural network model with oxidation stability as output variable was established.For the first time,the viscosity index is used as the input parameter of the prediction model,which greatly improves the prediction accuracy of the model.By analyzing the factors affecting the oxidation stability of hydrogenated base oil,the alkanes composition ofⅡ/Ⅲbase oil which has a positive/negative correlation with oxidation stability was found.InⅡ/Ⅲbase oils,the oxidation stability of lubricating oils increases with the increase of linear alkanes and the decrease of bicyclic and tricyclic alkanes when the aromatics content is very low.

关 键 词：润滑油 基础油 氧化安定性 神经网络 人工智能模拟 

分 类 号：TE6[石油与天然气工程—油气加工工程]