文献类型：期刊文献

英文题名：Enhancing Quality of Multivariate Process Monitoring Based on Vine Copula and Active Learning Strategy

作者：Zhou, Yang[1];Ren, Xiang[2];Li, Shaojun[1]

机构：[1]East China Univ Sci & Technol, Minist Educ, Key Lab Adv Control & Optimizat Chem Proc, Shanghai 200237, Peoples R China;[2]Rutgers State Univ, Dept Chem & Biochem Engn, Piscataway, NJ 08854 USA

年份：2018

卷号：57

期号：23

起止页码：7961

外文期刊名：INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH

收录：;EI(收录号：20182205258083);WOS:【SCI-EXPANDED(收录号:WOS:000435525200025)】；

基金：The authors of this paper appreciate the National Natural Science Foundation of China (under Project No. 21676086 and No. 21406064) and the Fundamental Research Funds for the Central Universities under Grant 222201717006 for their financial support.

语种：英文

外文关键词：Bayesian networks - Learning systems - Fault detection - Inference engines - Chemical detection - Process control

摘要：This paper proposes a new process monitoring method based on vine copula and active learning strategy under a limited number of labeled samples. The proposed method uses active learning strategy and the generalized Bayesian inference based probability (GBIP) index to choose samples that can provide the most significant information for the process monitoring model. An adaptive strategy is used to select the number of training samples in every active learning loop. Then, the vine copula-based dependence description (VCDD) is used to fulfill fault detection for complex chemical processes. The validity and effectiveness of the proposed approach are illustrated using a numerical example and the Tennessee Eastman (TE) benchmark process. The results show that the proposed method can maximize the process monitoring performance while minimizing the number of samples labeled.