石化企业循环冷却水系统腐蚀结垢预测模型的研究  

RESEARCH ON CORROSION AND SCALING PREDICTION MODLE FOR RECYCLE-COOLING WATER SYSTEM IN PETROCHEMICAL ENTERPRISES

作　　者：翁新龙 焦云强[2] 欧阳福生[1] 王建平[2] 邸雪梅[2] Weng Xinlong;Jiao Yunqiang;Ouyang Fusheng;Wang Jianping;Di Xuemei(Research Institute of Petroleum Processing,School of Chemical Engineering,East China University of Science and Technology,Shanghai 200237;Petro-Cyber Works Information Technology Co.,Ltd.)

机构地区：[1]华东理工大学化工学院石油加工研究所,上海200237 [2]石化盈科信息技术有限责任公司

出　　处：《石油炼制与化工》2023年第12期119-126,共8页Petroleum Processing and Petrochemicals

摘　　要：以某石化企业循环冷却水系统的运行数据为基础,通过预处理获得了899组有效数据样本;采用最大互信息系数和Pearson相关系数法,筛选出针对目标变量腐蚀速率(FSSL)和黏附速率(NFSL)预测模型的输入变量;分别运用BP神经网络、KNN回归和XGBoost机器学习算法建立了循环冷却水系统的FSSL和NFSL预测模型。对3种模型进行预测精准度和预警效果评价结果表明:3种模型的预测平均相对误差(MAPE)均在9%以下,都具备较好的拟合效果和泛化能力;其中基于XGBoost方法所建模型的性能最佳,其对FSSL和NFSL的MAPE均在5%以下,决定系数R 2均大于0.9,预警准确率分别在91.5%和97.3%以上。Based on the operating data from the cycle-cooling water system of a petrochemical enterprise,899 sets of valid data samples were obtained by data preprocessing;the input variables for corrosion rate prediction(FSSL)model and adhesion rate prediction(NFSL)model were selected by using maximum mutual information coefficient and Pearson correlation coefficient methods.With 3 machine learning algorithms including BP neural network,KNN regression,and XGBoost,the prediction models for FSSL and NFSL of the cycle-cooling water system were established respectively.The evaluation results of the prediction accuracy and warning effectiveness of three models showed that the average relative error of the all three models was below 9%,and the three models had good fitting effect and generalization ability.The model based on XGBoost method had the best performance,the average relative error for both FSSL and NFSL was less than 5%,the decision coefficient R 2 was over 0.9,and the early warning accuracy was over 91.5%and 97.3%respectively.

关 键 词：循环冷却水系统 腐蚀速率 黏附速率 机器学习算法 

分 类 号：TE65[石油与天然气工程—油气加工工程] TQ050.9[化学工程]