Optimal experiment design techniques integrated with time-series segmentation

L. Dobos, Z. Bankó, J. Abonyi

Research output: Conference contribution

Abstract

Process models play important role in computer aided process engineering. Although the structure of these models is a priori known, model parameters should be estimated based on experiments. The accuracy of the estimated parameters largely depends on the information content of the experimental data presented to the parameter identification algorithm. Optimal Experiment Design (OED) can maximize the confidence on the model parameters. Considering that OED is an iterative process, it may happen that the designed experiment contains segments which are not or less useful for parameter identification. Using the tools of the OED there is the opportunity to qualify the segments of the time-series of different data sets. After the segmentation, it will be possible to choose the most appropriate segments for identification of each parameter, i.e. to determine the parameters as accurate as possible.

Original languageEnglish
Title of host publicationSAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings
Pages207-210
Number of pages4
DOIs
Publication statusPublished - 2010
Event8th International Symposium on Applied Machine Intelligence and Informatics, SAMI 2010 - Herlany, Slovakia
Duration: jan. 28 2010jan. 30 2010

Other

Other8th International Symposium on Applied Machine Intelligence and Informatics, SAMI 2010
CountrySlovakia
CityHerlany
Period1/28/101/30/10

Fingerprint

Time series
Experiments
Identification (control systems)
Computer aided engineering
Process engineering

ASJC Scopus subject areas

  • Artificial Intelligence
  • Human-Computer Interaction
  • Information Systems

Cite this

Dobos, L., Bankó, Z., & Abonyi, J. (2010). Optimal experiment design techniques integrated with time-series segmentation. In SAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings (pp. 207-210). [5423737] https://doi.org/10.1109/SAMI.2010.5423737

Optimal experiment design techniques integrated with time-series segmentation. / Dobos, L.; Bankó, Z.; Abonyi, J.

SAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings. 2010. p. 207-210 5423737.

Research output: Conference contribution

Dobos, L, Bankó, Z & Abonyi, J 2010, Optimal experiment design techniques integrated with time-series segmentation. in SAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings., 5423737, pp. 207-210, 8th International Symposium on Applied Machine Intelligence and Informatics, SAMI 2010, Herlany, Slovakia, 1/28/10. https://doi.org/10.1109/SAMI.2010.5423737
Dobos L, Bankó Z, Abonyi J. Optimal experiment design techniques integrated with time-series segmentation. In SAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings. 2010. p. 207-210. 5423737 https://doi.org/10.1109/SAMI.2010.5423737
Dobos, L. ; Bankó, Z. ; Abonyi, J. / Optimal experiment design techniques integrated with time-series segmentation. SAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings. 2010. pp. 207-210
@inproceedings{b585f45f500642509b07c0f448153bfa,
title = "Optimal experiment design techniques integrated with time-series segmentation",
abstract = "Process models play important role in computer aided process engineering. Although the structure of these models is a priori known, model parameters should be estimated based on experiments. The accuracy of the estimated parameters largely depends on the information content of the experimental data presented to the parameter identification algorithm. Optimal Experiment Design (OED) can maximize the confidence on the model parameters. Considering that OED is an iterative process, it may happen that the designed experiment contains segments which are not or less useful for parameter identification. Using the tools of the OED there is the opportunity to qualify the segments of the time-series of different data sets. After the segmentation, it will be possible to choose the most appropriate segments for identification of each parameter, i.e. to determine the parameters as accurate as possible.",
keywords = "Optimal experiment design, Parameter identification, Segmentation, Time series",
author = "L. Dobos and Z. Bank{\'o} and J. Abonyi",
year = "2010",
doi = "10.1109/SAMI.2010.5423737",
language = "English",
isbn = "9781424464241",
pages = "207--210",
booktitle = "SAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings",

}

TY - GEN

T1 - Optimal experiment design techniques integrated with time-series segmentation

AU - Dobos, L.

AU - Bankó, Z.

AU - Abonyi, J.

PY - 2010

Y1 - 2010

N2 - Process models play important role in computer aided process engineering. Although the structure of these models is a priori known, model parameters should be estimated based on experiments. The accuracy of the estimated parameters largely depends on the information content of the experimental data presented to the parameter identification algorithm. Optimal Experiment Design (OED) can maximize the confidence on the model parameters. Considering that OED is an iterative process, it may happen that the designed experiment contains segments which are not or less useful for parameter identification. Using the tools of the OED there is the opportunity to qualify the segments of the time-series of different data sets. After the segmentation, it will be possible to choose the most appropriate segments for identification of each parameter, i.e. to determine the parameters as accurate as possible.

AB - Process models play important role in computer aided process engineering. Although the structure of these models is a priori known, model parameters should be estimated based on experiments. The accuracy of the estimated parameters largely depends on the information content of the experimental data presented to the parameter identification algorithm. Optimal Experiment Design (OED) can maximize the confidence on the model parameters. Considering that OED is an iterative process, it may happen that the designed experiment contains segments which are not or less useful for parameter identification. Using the tools of the OED there is the opportunity to qualify the segments of the time-series of different data sets. After the segmentation, it will be possible to choose the most appropriate segments for identification of each parameter, i.e. to determine the parameters as accurate as possible.

KW - Optimal experiment design

KW - Parameter identification

KW - Segmentation

KW - Time series

UR - http://www.scopus.com/inward/record.url?scp=77951428326&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951428326&partnerID=8YFLogxK

U2 - 10.1109/SAMI.2010.5423737

DO - 10.1109/SAMI.2010.5423737

M3 - Conference contribution

AN - SCOPUS:77951428326

SN - 9781424464241

SP - 207

EP - 210

BT - SAMI 2010 - 8th International Symposium on Applied Machine Intelligence and Informatics, Proceedings

ER -