PLS-based Similarity Analysis for Mode Identification in Multimode Manufacturing Processes

Ying ZHENG, S. Joe QIN*, Fu-Li WANG

*Corresponding author for this work

Research output: Journal PublicationsJournal Article (refereed)peer-review

11 Citations (Scopus)


Many industrial manufacturing processes have multiple operation modes because of different strategy and varying feedstock. The traditional statistical process monitoring tools such as PCA and PLS cannot be applied since they assume that the process must have single mode operation region only. In this paper, all the factors that will affect the change of the mode are considered, a similarity factor including the similarity factor of PLS models and the mean shift of the external variables is introduced to measure the similarity of two sets of data. On basis of this similarity factor, a moving window is used and a mode identification approach for multimode process monitoring is proposed. The proposed approach is demonstrated on the benchmark Tennessee Eastman process.
Original languageEnglish
Pages (from-to)777-782
Number of pages6
Issue number8
Early online date25 Sept 2015
Publication statusPublished - 2015
Externally publishedYes
Event9th IFAC Symposium on Advanced Control of Chemical Processes, ADCHEM 2015 - , Canada
Duration: 7 Jun 201510 Jun 2015

Bibliographical note

The authors are grateful for the financial support for this work from National Natural Science Foundation of China (61374139 and 61490700).


  • External analysis
  • Mode identification
  • Multiple operation mode
  • Partial least square(PLS)
  • Process monitoring


Dive into the research topics of 'PLS-based Similarity Analysis for Mode Identification in Multimode Manufacturing Processes'. Together they form a unique fingerprint.

Cite this