Title
MPCA based phase identification method and its application to process monitoring
Abstract
In order to characterize the intrinsic performance of multi-phase batch process further, a sub-phase partition method is proposed. According to the different numbers of principal components and variation direction of variable information, a two-step phase partition is realized for the phase partition of multi-phase process. After the two-step division, the entire time-slice matrices in the same sub-phase have the same number of principal components and similar variable variation direction. And the `fake' phases, stable phases and transition phases are identified by combining the specific characteristics of batch processes. The proposed MPCA modeling methods and steps based on sub-phase partition are given and applied to online monitoring of penicillin fermentation process.
Year
DOI
Venue
2012
10.1109/CDC.2012.6426888
CDC
Keywords
Field
DocType
process monitoring,mpca based phase identification method,time slice matrices,penicillin fermentation process,subphase partition method,fermentation,mpca modeling method,online monitoring,intrinsic performance,multiphase batch process,principal component analysis,multiphase process
Process engineering,Mathematical optimization,Biological system,Matrix (mathematics),Computer science,Batch processing,Partition (number theory),Principal component analysis,Partition method
Conference
Volume
Issue
ISSN
null
null
0743-1546 E-ISBN : 978-1-4673-2064-1
ISBN
Citations 
PageRank 
978-1-4673-2064-1
0
0.34
References 
Authors
0
5
Name
Order
Citations
PageRank
Yuqing Chang101.35
Shu Wang200.34
Shuai Tan300.34
Fuli Wang45212.61
Zhi-Zhong Mao5306.07