Exact State Reconstruction for LTI-Systems with Non-Differentiable Unknown Inputs

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

In this paper, a sliding mode based observer design for strongly observable systems in the presence of non-differentiable unknown inputs is presented. The system state is reconstructed in finite time by means of a linear observer in combination with a robust exact differentiator. By appropriate filtering of the measured output, differentiation of signal components involving the unknown input is avoided. For strongly detectable systems, the proposed technique is combined with invariant subspace methods to provide asymptotic reconstruction of the state. A numerical simulation example demonstrates the performance of the proposed approach.
Original languageEnglish
Title of host publication18th European Control Conference (ECC)
Pages3096-3102
ISBN (Electronic)978-3-907144-01-5
DOIs
Publication statusPublished - 2019
Event18th European Control Conference - Naples, Italy
Duration: 25 Jun 201928 Jun 2019

Conference

Conference18th European Control Conference
Abbreviated titleECC
CountryItaly
CityNaples
Period25/06/1928/06/19

Fingerprint

Computer simulation

Cite this

Exact State Reconstruction for LTI-Systems with Non-Differentiable Unknown Inputs. / Tranninger, Markus; Seeber, Richard; Steinberger, Martin; Horn, Martin.

18th European Control Conference (ECC). 2019. p. 3096-3102.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Tranninger, M, Seeber, R, Steinberger, M & Horn, M 2019, Exact State Reconstruction for LTI-Systems with Non-Differentiable Unknown Inputs. in 18th European Control Conference (ECC). pp. 3096-3102, 18th European Control Conference, Naples, Italy, 25/06/19. https://doi.org/10.23919/ECC.2019.8796142
@inproceedings{835c479426b14f8bba3a8d6c0267f8ad,
title = "Exact State Reconstruction for LTI-Systems with Non-Differentiable Unknown Inputs",
abstract = "In this paper, a sliding mode based observer design for strongly observable systems in the presence of non-differentiable unknown inputs is presented. The system state is reconstructed in finite time by means of a linear observer in combination with a robust exact differentiator. By appropriate filtering of the measured output, differentiation of signal components involving the unknown input is avoided. For strongly detectable systems, the proposed technique is combined with invariant subspace methods to provide asymptotic reconstruction of the state. A numerical simulation example demonstrates the performance of the proposed approach.",
author = "Markus Tranninger and Richard Seeber and Martin Steinberger and Martin Horn",
year = "2019",
doi = "10.23919/ECC.2019.8796142",
language = "English",
pages = "3096--3102",
booktitle = "18th European Control Conference (ECC)",

}

TY - GEN

T1 - Exact State Reconstruction for LTI-Systems with Non-Differentiable Unknown Inputs

AU - Tranninger, Markus

AU - Seeber, Richard

AU - Steinberger, Martin

AU - Horn, Martin

PY - 2019

Y1 - 2019

N2 - In this paper, a sliding mode based observer design for strongly observable systems in the presence of non-differentiable unknown inputs is presented. The system state is reconstructed in finite time by means of a linear observer in combination with a robust exact differentiator. By appropriate filtering of the measured output, differentiation of signal components involving the unknown input is avoided. For strongly detectable systems, the proposed technique is combined with invariant subspace methods to provide asymptotic reconstruction of the state. A numerical simulation example demonstrates the performance of the proposed approach.

AB - In this paper, a sliding mode based observer design for strongly observable systems in the presence of non-differentiable unknown inputs is presented. The system state is reconstructed in finite time by means of a linear observer in combination with a robust exact differentiator. By appropriate filtering of the measured output, differentiation of signal components involving the unknown input is avoided. For strongly detectable systems, the proposed technique is combined with invariant subspace methods to provide asymptotic reconstruction of the state. A numerical simulation example demonstrates the performance of the proposed approach.

U2 - 10.23919/ECC.2019.8796142

DO - 10.23919/ECC.2019.8796142

M3 - Conference contribution

SP - 3096

EP - 3102

BT - 18th European Control Conference (ECC)

ER -