Nonlinear system identification by means of SVMs: choice of excitation signals

Anna Marconato; Andrea Boni; D. Petri; Joannes Schoukens

Nonlinear system identification by means of SVMs: choice of excitation signals

Anna Marconato, Andrea Boni, D. Petri, Joannes Schoukens

Electricity

Research output: Chapter in Book/Report/Conference proceeding › Conference paper

3 Citations (Scopus)

Abstract

In this work we discuss the application of Support Vector Machines to the problem of identifying a specific class of nonlinear systems, namely Wiener-Hammerstein systems. Only based on a set of Input/Output measurements, a model is built that approximates well the behavior of the considered system. However, care should be taken when designing suitable excitations, as the performance of the proposed approach turns out to be quite sensitive to the nature of the input signal. This sensitivity is studied here by using several datasets, characterised by different excitation signals, in terms of root mean square value, frequency band, spectrum shape and amplitude distribution.

Original language	English
Title of host publication	16th IMEKO TC4. International Symposium, 2008, Florence, Italy September 22 - 24, 2008
Number of pages	6
Publication status	Published - 22 Sept 2008
Event	Finds and Results from the Swedish Cyprus Expedition: A Gender Perspective at the Medelhavsmuseet - Stockholm, Sweden Duration: 21 Sept 2009 → 25 Sept 2009

Conference

Conference	Finds and Results from the Swedish Cyprus Expedition: A Gender Perspective at the Medelhavsmuseet
Country/Territory	Sweden
City	Stockholm
Period	21/09/09 → 25/09/09

Keywords

Support Vector Machines
SVMs
Input/Output measurements

Cite this

@inproceedings{5cd49d3142d14f0683666494bf87e184,

title = "Nonlinear system identification by means of SVMs: choice of excitation signals",

abstract = "In this work we discuss the application of Support Vector Machines to the problem of identifying a specific class of nonlinear systems, namely Wiener-Hammerstein systems. Only based on a set of Input/Output measurements, a model is built that approximates well the behavior of the considered system. However, care should be taken when designing suitable excitations, as the performance of the proposed approach turns out to be quite sensitive to the nature of the input signal. This sensitivity is studied here by using several datasets, characterised by different excitation signals, in terms of root mean square value, frequency band, spectrum shape and amplitude distribution.",

keywords = "Support Vector Machines, SVMs, Input/Output measurements",

author = "Anna Marconato and Andrea Boni and D. Petri and Joannes Schoukens",

year = "2008",

month = sep,

day = "22",

language = "English",

booktitle = "16th IMEKO TC4. International Symposium, 2008, Florence, Italy September 22 - 24, 2008",

note = "Finds and Results from the Swedish Cyprus Expedition: A Gender Perspective at the Medelhavsmuseet ; Conference date: 21-09-2009 Through 25-09-2009",

}

TY - GEN

T1 - Nonlinear system identification by means of SVMs: choice of excitation signals

AU - Marconato, Anna

AU - Boni, Andrea

AU - Petri, D.

AU - Schoukens, Joannes

PY - 2008/9/22

Y1 - 2008/9/22

N2 - In this work we discuss the application of Support Vector Machines to the problem of identifying a specific class of nonlinear systems, namely Wiener-Hammerstein systems. Only based on a set of Input/Output measurements, a model is built that approximates well the behavior of the considered system. However, care should be taken when designing suitable excitations, as the performance of the proposed approach turns out to be quite sensitive to the nature of the input signal. This sensitivity is studied here by using several datasets, characterised by different excitation signals, in terms of root mean square value, frequency band, spectrum shape and amplitude distribution.

AB - In this work we discuss the application of Support Vector Machines to the problem of identifying a specific class of nonlinear systems, namely Wiener-Hammerstein systems. Only based on a set of Input/Output measurements, a model is built that approximates well the behavior of the considered system. However, care should be taken when designing suitable excitations, as the performance of the proposed approach turns out to be quite sensitive to the nature of the input signal. This sensitivity is studied here by using several datasets, characterised by different excitation signals, in terms of root mean square value, frequency band, spectrum shape and amplitude distribution.

KW - Support Vector Machines

KW - SVMs

KW - Input/Output measurements

M3 - Conference paper

BT - 16th IMEKO TC4. International Symposium, 2008, Florence, Italy September 22 - 24, 2008

T2 - Finds and Results from the Swedish Cyprus Expedition: A Gender Perspective at the Medelhavsmuseet

Y2 - 21 September 2009 through 25 September 2009

ER -

Nonlinear system identification by means of SVMs: choice of excitation signals

Abstract

Conference

Keywords

Fingerprint

Cite this