The Decoupled Polynomial NARX Model: Parameter Reduction and Structural Insights for the Bouc-Wen Benchmark

David Westwick; Gabriel Hollander; Joannes Schoukens

The Decoupled Polynomial NARX Model: Parameter Reduction and Structural Insights for the Bouc-Wen Benchmark

David Westwick, Gabriel Hollander, Joannes Schoukens

Onderzoeksoutput: Unpublished paper

Samenvatting

We generated identification data from the default set-up of the Bouc-Wen Benchmark [3], and identified the 7 models listed in Table 1. All models used 8 past outputs and 8 past inputs. The sigmoidnet NARX model included a direct feedthrough term. All polynomial models were of degree 3. The models were then tested in both simulation and onestep-ahead prediction on the two test signals provided with the benchmark.

Originele taal-2	English
Pagina's	44
Aantal pagina's	1
Status	Published - 24 apr. 2017
Evenement	Workshop on Nonlinear System Identification Benchmarks 2017 - Vrije Universiteit Brussels, U-residence, Brussels, Belgium Duur: 24 apr. 2017 → 26 apr. 2017 http://homepages.vub.ac.be/~mschouke/FILES/BenchmarkWorkshop2017_Abstracts.pdf

Conference

Conference	Workshop on Nonlinear System Identification Benchmarks 2017
Land/Regio	Belgium
Stad	Brussels
Periode	24/04/17 → 26/04/17
Internet adres	http://homepages.vub.ac.be/~mschouke/FILES/BenchmarkWorkshop2017_Abstracts.pdf

Toegang tot document

http://homepages.vub.ac.be/~mschouke/FILES/BenchmarkWorkshop2017_Abstracts.pdf

Citeer dit

@conference{911bc308283148eb91084d8994604162,

title = "The Decoupled Polynomial NARX Model: Parameter Reduction and Structural Insights for the Bouc-Wen Benchmark",

abstract = "We generated identification data from the default set-up of the Bouc-Wen Benchmark [3], and identified the 7 models listed in Table 1. All models used 8 past outputs and 8 past inputs. The sigmoidnet NARX model included a direct feedthrough term. All polynomial models were of degree 3. The models were then tested in both simulation and onestep-ahead prediction on the two test signals provided with the benchmark.",

author = "David Westwick and Gabriel Hollander and Joannes Schoukens",

year = "2017",

month = apr,

day = "24",

language = "English",

pages = "44",

note = "Workshop on Nonlinear System Identification Benchmarks 2017 ; Conference date: 24-04-2017 Through 26-04-2017",

url = "http://homepages.vub.ac.be/~mschouke/FILES/BenchmarkWorkshop2017_Abstracts.pdf",

}

Westwick, D, Hollander, G & Schoukens, J 2017, 'The Decoupled Polynomial NARX Model: Parameter Reduction and Structural Insights for the Bouc-Wen Benchmark', Paper gepresenteerd op Workshop on Nonlinear System Identification Benchmarks 2017, Brussels, Belgium, 24/04/17 - 26/04/17 blz. 44. <http://homepages.vub.ac.be/~mschouke/FILES/BenchmarkWorkshop2017_Abstracts.pdf>

TY - CONF

T1 - The Decoupled Polynomial NARX Model: Parameter Reduction and Structural Insights for the Bouc-Wen Benchmark

AU - Westwick, David

AU - Hollander, Gabriel

AU - Schoukens, Joannes

PY - 2017/4/24

Y1 - 2017/4/24

N2 - We generated identification data from the default set-up of the Bouc-Wen Benchmark [3], and identified the 7 models listed in Table 1. All models used 8 past outputs and 8 past inputs. The sigmoidnet NARX model included a direct feedthrough term. All polynomial models were of degree 3. The models were then tested in both simulation and onestep-ahead prediction on the two test signals provided with the benchmark.

AB - We generated identification data from the default set-up of the Bouc-Wen Benchmark [3], and identified the 7 models listed in Table 1. All models used 8 past outputs and 8 past inputs. The sigmoidnet NARX model included a direct feedthrough term. All polynomial models were of degree 3. The models were then tested in both simulation and onestep-ahead prediction on the two test signals provided with the benchmark.

M3 - Unpublished paper

SP - 44

T2 - Workshop on Nonlinear System Identification Benchmarks 2017

Y2 - 24 April 2017 through 26 April 2017

ER -

The Decoupled Polynomial NARX Model: Parameter Reduction and Structural Insights for the Bouc-Wen Benchmark

Samenvatting

Conference

Toegang tot document

Vingerafdruk

Citeer dit