Fisher information matrix based time-series segmentation of process data

László Dobos, J. Abonyi

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Advanced chemical process engineering tools, like model predictive control or soft sensor solutions require proper process models. Parameter identification of these models needs input-output data with high information content. When model based optimal experimental design techniqes cannot be applied, the extraction of informative segements from historical data can also support system identification. We developed a goal-oriented Fisher information based time-series segmentation algorithm, aimed at selecting informative segments from historical process data. The utilized standard bottom-up algorithm is widely used in off-line analysis of process data. Different segments can support the identification of parameter sets. Hence, instead of using either D- or E-optimality as the criterion for comparing the information content of two input sequences (neigbouring segments), we propose the use of Krzanowski's similarity coefficient between the eigenvectors of the Fisher information matrices obtained from the sequences. The efficiency of the proposed methodology is demonstrated by two application examples. The algorithm is capable to extract segments with parameter-set specific information content from historical process data.

Original languageEnglish
Pages (from-to)99-108
Number of pages10
JournalChemical Engineering Science
Volume101
DOIs
Publication statusPublished - Sep 20 2013

Fingerprint

Fisher information matrix
Fisher Information Matrix
Time series
Information Content
Segmentation
Identification (control systems)
E-optimality
Model predictive control
Process engineering
Chemical engineering
Similarity Coefficient
Optimal Experimental Design
Soft Sensor
D-optimality
Eigenvalues and eigenfunctions
Design of experiments
Fisher Information
Chemical Processes
Historical Data
Model Predictive Control

Keywords

  • Fisher information matrix
  • Optimal experimental design (OED)
  • Parameter identification
  • Process model
  • Time-series segmentation

ASJC Scopus subject areas

  • Chemical Engineering(all)
  • Chemistry(all)
  • Applied Mathematics
  • Industrial and Manufacturing Engineering

Cite this

Fisher information matrix based time-series segmentation of process data. / Dobos, László; Abonyi, J.

In: Chemical Engineering Science, Vol. 101, 20.09.2013, p. 99-108.

Research output: Contribution to journalArticle

@article{dd3f45400e5b4d59a7b59c2c51df81d3,
title = "Fisher information matrix based time-series segmentation of process data",
abstract = "Advanced chemical process engineering tools, like model predictive control or soft sensor solutions require proper process models. Parameter identification of these models needs input-output data with high information content. When model based optimal experimental design techniqes cannot be applied, the extraction of informative segements from historical data can also support system identification. We developed a goal-oriented Fisher information based time-series segmentation algorithm, aimed at selecting informative segments from historical process data. The utilized standard bottom-up algorithm is widely used in off-line analysis of process data. Different segments can support the identification of parameter sets. Hence, instead of using either D- or E-optimality as the criterion for comparing the information content of two input sequences (neigbouring segments), we propose the use of Krzanowski's similarity coefficient between the eigenvectors of the Fisher information matrices obtained from the sequences. The efficiency of the proposed methodology is demonstrated by two application examples. The algorithm is capable to extract segments with parameter-set specific information content from historical process data.",
keywords = "Fisher information matrix, Optimal experimental design (OED), Parameter identification, Process model, Time-series segmentation",
author = "L{\'a}szl{\'o} Dobos and J. Abonyi",
year = "2013",
month = "9",
day = "20",
doi = "10.1016/j.ces.2013.06.030",
language = "English",
volume = "101",
pages = "99--108",
journal = "Chemical Engineering Science",
issn = "0009-2509",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - Fisher information matrix based time-series segmentation of process data

AU - Dobos, László

AU - Abonyi, J.

PY - 2013/9/20

Y1 - 2013/9/20

N2 - Advanced chemical process engineering tools, like model predictive control or soft sensor solutions require proper process models. Parameter identification of these models needs input-output data with high information content. When model based optimal experimental design techniqes cannot be applied, the extraction of informative segements from historical data can also support system identification. We developed a goal-oriented Fisher information based time-series segmentation algorithm, aimed at selecting informative segments from historical process data. The utilized standard bottom-up algorithm is widely used in off-line analysis of process data. Different segments can support the identification of parameter sets. Hence, instead of using either D- or E-optimality as the criterion for comparing the information content of two input sequences (neigbouring segments), we propose the use of Krzanowski's similarity coefficient between the eigenvectors of the Fisher information matrices obtained from the sequences. The efficiency of the proposed methodology is demonstrated by two application examples. The algorithm is capable to extract segments with parameter-set specific information content from historical process data.

AB - Advanced chemical process engineering tools, like model predictive control or soft sensor solutions require proper process models. Parameter identification of these models needs input-output data with high information content. When model based optimal experimental design techniqes cannot be applied, the extraction of informative segements from historical data can also support system identification. We developed a goal-oriented Fisher information based time-series segmentation algorithm, aimed at selecting informative segments from historical process data. The utilized standard bottom-up algorithm is widely used in off-line analysis of process data. Different segments can support the identification of parameter sets. Hence, instead of using either D- or E-optimality as the criterion for comparing the information content of two input sequences (neigbouring segments), we propose the use of Krzanowski's similarity coefficient between the eigenvectors of the Fisher information matrices obtained from the sequences. The efficiency of the proposed methodology is demonstrated by two application examples. The algorithm is capable to extract segments with parameter-set specific information content from historical process data.

KW - Fisher information matrix

KW - Optimal experimental design (OED)

KW - Parameter identification

KW - Process model

KW - Time-series segmentation

UR - http://www.scopus.com/inward/record.url?scp=84880379242&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84880379242&partnerID=8YFLogxK

U2 - 10.1016/j.ces.2013.06.030

DO - 10.1016/j.ces.2013.06.030

M3 - Article

AN - SCOPUS:84880379242

VL - 101

SP - 99

EP - 108

JO - Chemical Engineering Science

JF - Chemical Engineering Science

SN - 0009-2509

ER -