基于大数据的电脱盐系统智能优化与影响因素分析  

Intelligent optimization and analysis of an electro-desalting system based on big data

作　　者：朱建新 周杨 乔松 袁文彬 方向荣 亢海洲 吕宝林 ZHU Jian-xin;ZHOU Yang;QIAO Song;YUAN Wen-bin;FANG Xiang-rong;KANG Hai-zhou;LYU Bao-lin(Hefei General Machinery Research Institute Co.Ltd.,Hefei 230031,China;National Technology Research Center for Safety Engineering of Pressure Vessels and Pipelines,Hefei 230031,China)

机构地区：[1]合肥通用机械研究院有限公司,安徽合肥230031 [2]国家压力容器与管道安全工程技术研究中心,安徽合肥230031

出　　处：《高校化学工程学报》2023年第1期95-102,共8页Journal of Chemical Engineering of Chinese Universities

基　　金：国家重点研发计划(2022YFB3306401);安徽省科技重大专项(17030901014);国机集团重大科技专项项目(国机科[2017]456号);安徽省自然科学基金(2008085J24)。

摘　　要：针对电脱盐系统影响因素耦合关联,脱盐后盐的质量浓度控制困难等问题,提出基于贝叶斯分类的电脱盐系统脱盐后盐的质量浓度智能预测模型,建立基于模型及线性规划的脱盐系统优化方法。利用从分布式控制系统(DCS)和实验室信息管理系统(LIMS)中获取的运行数据,筛选影响电脱盐效果的主要因素,建立基于40个运行参数的脱盐后盐的质量浓度智能监测模型。模型对脱盐后盐的质量浓度1.5~2.0、2.0~3.0 mg·L^(-1)和大于3.0 mg·L^(-1)的识别召回率分别达到89.6%、84.4%和70.0%。利用优化方法识别了脱盐罐油水界位、操作温度、混合压差及注水量等影响脱盐效果的关键因素,通过对电脱盐操作参数进行优化,显著降低了脱盐后盐的质量浓度,实现系统的优化运行。The performance of electro-desalting systems are affected by multiple factors,which makes it difficult in maintaining stable desalting performance.An intelligent prediction model for after-desalination salt contents based on Bayes classifier was proposed.An optimization method based on modeling and linear programming was established,and an after-desalination salt content prediction model based on 40 operating parameters was obtained using operation data obtained from the distributed control system (DCS) and the laboratory information management system (LIMS).For samples with after-desalination salt contents range from 1.5-2.0,2.0-3.0 and>3.0 mg·L^(-1),the model recall rates are 89.6%,84.4%and 70.0%,respectively.Factors including level of water-oil boundary in desalting vessel,operating temperature,mixing pressure drop of oil and demulsifier and the amount of water injected were studied.The after-desalination salt content can be significantly reduced under optimized operation following optimization suggestions.

关 键 词：电脱盐 大数据 智能监测 影响因素 优化 

分 类 号：TQ465.92[化学工程—制药化工]