Abstract

Multiscale wavelet-based representation is a powerful data analysis and feature extraction tool. In this paper, this characteristic of multiscale representation is utilized to improve the prediction accuracy of nonlinear models by developing a multiscale nonlinear (MSNL) system identification algorithm. In particular, we consider the class of linear-in-the-parameters nonlinear models with known basis functions. The idea is to decompose the input-output data, construct multiple nonlinear models at multiple scales using the scaled signal approximations of the data, and then select among all MSNL models the one which best describes the process. The main advantage of the MSNL modeling algorithm is that it inherently accounts for the presence of noise in the data by the application of low pass Alters used in the multiscale decomposition, which in turn improves the model robustness to measurement noise in the data and thus enhances its prediction. This advantage of MSNL modeling is demonstrated using a reactor model with nonlinear reaction rate.

Original languageEnglish
Title of host publicationProceedings of the 46th IEEE Conference on Decision and Control 2007, CDC
Pages5755-5760
Number of pages6
DOIs
Publication statusPublished - 2007
Event46th IEEE Conference on Decision and Control 2007, CDC - New Orleans, LA, United States
Duration: 12 Dec 200714 Dec 2007

Other

Other46th IEEE Conference on Decision and Control 2007, CDC
CountryUnited States
CityNew Orleans, LA
Period12/12/0714/12/07

Fingerprint

Nonlinear System Identification
Nonlinear Model
Nonlinear systems
Identification (control systems)
Multiscale Modeling
Nonlinear Modeling
Model Robustness
Decompose
Multiscale Model
Multiple Scales
Prediction
Multiple Models
Reaction Rate
Reactor
Feature Extraction
Basis Functions
Data analysis
Wavelets
Reaction rates
Feature extraction

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Modelling and Simulation
  • Control and Optimization

Cite this

Nounou, M. N., & Nounou, H. N. (2007). Multiscale nonlinear system identification. In Proceedings of the 46th IEEE Conference on Decision and Control 2007, CDC (pp. 5755-5760). [4434311] https://doi.org/10.1109/CDC.2007.4434311

Multiscale nonlinear system identification. / Nounou, Mohamed N.; Nounou, Hazem N.

Proceedings of the 46th IEEE Conference on Decision and Control 2007, CDC. 2007. p. 5755-5760 4434311.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nounou, MN & Nounou, HN 2007, Multiscale nonlinear system identification. in Proceedings of the 46th IEEE Conference on Decision and Control 2007, CDC., 4434311, pp. 5755-5760, 46th IEEE Conference on Decision and Control 2007, CDC, New Orleans, LA, United States, 12/12/07. https://doi.org/10.1109/CDC.2007.4434311
Nounou MN, Nounou HN. Multiscale nonlinear system identification. In Proceedings of the 46th IEEE Conference on Decision and Control 2007, CDC. 2007. p. 5755-5760. 4434311 https://doi.org/10.1109/CDC.2007.4434311
Nounou, Mohamed N. ; Nounou, Hazem N. / Multiscale nonlinear system identification. Proceedings of the 46th IEEE Conference on Decision and Control 2007, CDC. 2007. pp. 5755-5760
@inproceedings{cfbe4fa9ec784d3597a11a1de02e170e,
title = "Multiscale nonlinear system identification",
abstract = "Multiscale wavelet-based representation is a powerful data analysis and feature extraction tool. In this paper, this characteristic of multiscale representation is utilized to improve the prediction accuracy of nonlinear models by developing a multiscale nonlinear (MSNL) system identification algorithm. In particular, we consider the class of linear-in-the-parameters nonlinear models with known basis functions. The idea is to decompose the input-output data, construct multiple nonlinear models at multiple scales using the scaled signal approximations of the data, and then select among all MSNL models the one which best describes the process. The main advantage of the MSNL modeling algorithm is that it inherently accounts for the presence of noise in the data by the application of low pass Alters used in the multiscale decomposition, which in turn improves the model robustness to measurement noise in the data and thus enhances its prediction. This advantage of MSNL modeling is demonstrated using a reactor model with nonlinear reaction rate.",
author = "Nounou, {Mohamed N.} and Nounou, {Hazem N.}",
year = "2007",
doi = "10.1109/CDC.2007.4434311",
language = "English",
isbn = "1424414989",
pages = "5755--5760",
booktitle = "Proceedings of the 46th IEEE Conference on Decision and Control 2007, CDC",

}

TY - GEN

T1 - Multiscale nonlinear system identification

AU - Nounou, Mohamed N.

AU - Nounou, Hazem N.

PY - 2007

Y1 - 2007

N2 - Multiscale wavelet-based representation is a powerful data analysis and feature extraction tool. In this paper, this characteristic of multiscale representation is utilized to improve the prediction accuracy of nonlinear models by developing a multiscale nonlinear (MSNL) system identification algorithm. In particular, we consider the class of linear-in-the-parameters nonlinear models with known basis functions. The idea is to decompose the input-output data, construct multiple nonlinear models at multiple scales using the scaled signal approximations of the data, and then select among all MSNL models the one which best describes the process. The main advantage of the MSNL modeling algorithm is that it inherently accounts for the presence of noise in the data by the application of low pass Alters used in the multiscale decomposition, which in turn improves the model robustness to measurement noise in the data and thus enhances its prediction. This advantage of MSNL modeling is demonstrated using a reactor model with nonlinear reaction rate.

AB - Multiscale wavelet-based representation is a powerful data analysis and feature extraction tool. In this paper, this characteristic of multiscale representation is utilized to improve the prediction accuracy of nonlinear models by developing a multiscale nonlinear (MSNL) system identification algorithm. In particular, we consider the class of linear-in-the-parameters nonlinear models with known basis functions. The idea is to decompose the input-output data, construct multiple nonlinear models at multiple scales using the scaled signal approximations of the data, and then select among all MSNL models the one which best describes the process. The main advantage of the MSNL modeling algorithm is that it inherently accounts for the presence of noise in the data by the application of low pass Alters used in the multiscale decomposition, which in turn improves the model robustness to measurement noise in the data and thus enhances its prediction. This advantage of MSNL modeling is demonstrated using a reactor model with nonlinear reaction rate.

UR - http://www.scopus.com/inward/record.url?scp=62749096779&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=62749096779&partnerID=8YFLogxK

U2 - 10.1109/CDC.2007.4434311

DO - 10.1109/CDC.2007.4434311

M3 - Conference contribution

SN - 1424414989

SN - 9781424414987

SP - 5755

EP - 5760

BT - Proceedings of the 46th IEEE Conference on Decision and Control 2007, CDC

ER -