Profitability related industrial-scale batch processes monitoring via deep learning based soft sensor development

被引:29
|
作者
Ji, Cheng [1 ]
Ma, Fangyuan [1 ,2 ]
Wang, Jingde [1 ]
Sun, Wei [1 ]
机构
[1] Beijing Univ Chem Technol, Coll Chem Engn, North Third Ring Rd 15, Beijing 100029, Peoples R China
[2] Tsinghua Univ, Wuxi Res Inst Appl Technol, Ctr Proc Monitoring & Data Anal, Wuxi 214072, Peoples R China
基金
中国国家自然科学基金;
关键词
Differential recurrent neural network; Data stacked strategy; Variable selection; Mutual information; Product quality prediction and monitoring; Industrial penicillin fermentation process; NEURAL-NETWORKS; FAULT-DETECTION; REGRESSION; FERMENTATION;
D O I
10.1016/j.compchemeng.2022.108125
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Data-driven soft sensor technology has been widely developed to estimate quality-related variables, while following difficulties still limit its application in batch processes, such as different initial conditions, uneven-length of batches, and the extraction of within-batch multiphase features. To address these problems, a qual-ity prediction and monitoring framework is proposed. Variables related to quality-related variables are first selected, and a data stacked strategy is proposed to transform three-dimensional batch data into time-lagged sequences that can be fed into soft sensor models. Aiming to extract the multiphase features, a novel differen-tial recurrent neural networks is established by embedding differential operations into long short-term memory neural networks. Moreover, to ensure profitability, prediction residuals are employed for quality monitoring. Case study on a simulation dataset and an industrial-scale penicillin fermentation process demonstrates the effectiveness of the proposed method and its applicability to batch process monitoring and control in both ac-ademic research and industrial operation.
引用
收藏
页数:15
相关论文
empty
未找到相关数据