Online flooding monitoring in packed towers using EDPCA method

Author

Wenwen, Wang ; Zewen, Cao ; Zengliang, Gao ; Yi, Liu

Author_Institution

Engineering Research Center of Process Equipment and Remanufacturing, Ministry of Education, Institute of Process Equipment and Control Engineering, Zhejiang University of Technology, Hangzhou, 310014, PR China

fYear

2015

fDate

28-30 July 2015

Firstpage

8258

Lastpage

8263

Abstract

Traditional flooding monitoring methods have been found insufficient to monitor various types of industrial packed towers. In this work, an enhanced data-driven monitoring method, i.e., enhanced dynamic principal component analysis (EDPCA), is proposed for online flooding monitoring in packed towers. The operation data samples are first clustered into several classes using the fuzzy c-means clustering approach. Then, several single DPCA models are trained with each subset. Furthermore, the Bayesian inference is adopted to integrate these single DPCA models. The obtained results for online flooding monitoring of an air-water packed tower demonstrate that EDPCA can obtain better and more reliable performance, compared with the DPCA method.

Keywords

Floods; Liquids; Monitoring; Poles and towers; Principal component analysis; Silicon; Bayesian inference; Flooding monitoring; dynamic principal component analysis; packed towers;

fLanguage

English

Publisher

ieee

Conference_Titel

Control Conference (CCC), 2015 34th Chinese

Conference_Location

Hangzhou, China

Type

conf

DOI

10.1109/ChiCC.2015.7260950

Filename

7260950