基于主成分分析-支持向量机的电脱盐工艺预测方法研究  

Research on Prediction Method of Electric Desalination Process Based on Principal Component Analysis-Support Vector Machine

作　　者：乔松 吕宝林 朱建新 袁文彬 QIAO Song;LV Bao-lin;ZHU Jian-xin;YUAN Wen-bin(Hefei General Machinery Research Institute Co.,Ltd.;National Technical Research Center for Safety Engineering of Pressure Vessels and Pipelines)

机构地区：[1]合肥通用机械研究院有限公司 [2]国家压力容器与管道安全工程技术研究中心

出　　处：《化工自动化及仪表》2024年第3期443-448,共6页Control and Instruments in Chemical Industry

基　　金：国家重点研发计划(批准号:2022YFB3306400)资助的课题;安徽省重点研究与开发计划(批准号:202104a07020011)资助的课题。

摘　　要：原油电脱盐工艺水平直接影响常减压装置腐蚀状态和工艺催化剂活性。综合考虑电脱盐工艺参数,选择脱盐前盐质量浓度、含水量及电场强度等40个工艺参数共440组工况数据,进行电脱盐效果评价研究。根据脱后盐含量数据箱型图特征,将电脱盐效果等级分为3级,采用主成分分析法,提取原始数据累积贡献率90.91%以上的前3个主成分,与电脱盐等级建立数据集,最后通过支持向量机建立电脱盐等级评价模型。结果表明:该方法对指导电脱盐工艺优化具有较好的工程适用性。The technical level of crude oil's electro-desalting process directly influences the corrosion status of atmospheric and vacuum distillation unit and the activity of process catalyst.In this paper,40 process parameters including salt mass concentration,water content and electric field strength before desalting were selected to evaluate the electro-desalting effect,total 440 sets of data of working conditions.Through considering characteristics of the desalted salt content box diagram,the electro-desalting effect was divided into three levels,and the principal component analysis method was adopted to extract the first three principal components with the original data utilization rate of more than 90.91%,and the data set was established with the electro-desalting level.Finally,the evaluation model of the electro-desalting level was established by the support vector machine.The results show that,the method proposed has good engineering applicability in guiding optimization of the electro-desalting process.

关 键 词：主成分分析 支持向量机 电脱盐 工艺参数 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]