Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods

O. R. De Lautour, P. Omenzetter

Research output: Chapter in Book/Report/Conference proceedingChapter (peer-reviewed)

Abstract

Developed for analysing long, periodic records, time series analysis methods are inherently suited for Structural Health Monitoring (SHM) applications. In this research, Autoregressive (AR) models were used to fit acceleration time histories from a 3-storey bookshelf laboratory structure and the ASCE Phase II SHM Benchmark Structure in healthy and several damaged states. Preliminary visual inspection of the large sets of AR coefficients to check the presence of clusters corresponding to different damage severities was achieved using Sammon mapping. Systematic classification of damage into states based on the analysis of the AR coefficients was achieved using two supervised classification techniques: Nearest Neighbour (NN) and Learning Vector Quantisation (LVQ), and one unsupervised technique: Self-Organising Maps (SOM).
Original languageEnglish
Title of host publicationFutures in Mechanics of Structures and Materials
Subtitle of host publicationProceedings of the 20th Australasian Conference on the Mechanics of Structures and Materials, ACMSM20
EditorsThiru Aravinthan, Warna Karunasena, Hao Wang
PublisherCRC Press
Pages123-129
Number of pages7
ISBN (Print)9780415491969
DOIs
Publication statusPublished - 20 Nov 2008

Fingerprint

Time series analysis
Structural health monitoring
Vector quantization
Self organizing maps
Inspection

Cite this

De Lautour, O. R., & Omenzetter, P. (2008). Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods. In T. Aravinthan, W. Karunasena, & H. Wang (Eds.), Futures in Mechanics of Structures and Materials: Proceedings of the 20th Australasian Conference on the Mechanics of Structures and Materials, ACMSM20 (pp. 123-129). CRC Press. https://doi.org/10.13140/2.1.3574.4329

Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods. / De Lautour, O. R.; Omenzetter, P.

Futures in Mechanics of Structures and Materials: Proceedings of the 20th Australasian Conference on the Mechanics of Structures and Materials, ACMSM20. ed. / Thiru Aravinthan; Warna Karunasena; Hao Wang. CRC Press, 2008. p. 123-129.

Research output: Chapter in Book/Report/Conference proceedingChapter (peer-reviewed)

De Lautour, OR & Omenzetter, P 2008, Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods. in T Aravinthan, W Karunasena & H Wang (eds), Futures in Mechanics of Structures and Materials: Proceedings of the 20th Australasian Conference on the Mechanics of Structures and Materials, ACMSM20. CRC Press, pp. 123-129. https://doi.org/10.13140/2.1.3574.4329
De Lautour OR, Omenzetter P. Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods. In Aravinthan T, Karunasena W, Wang H, editors, Futures in Mechanics of Structures and Materials: Proceedings of the 20th Australasian Conference on the Mechanics of Structures and Materials, ACMSM20. CRC Press. 2008. p. 123-129 https://doi.org/10.13140/2.1.3574.4329
De Lautour, O. R. ; Omenzetter, P. / Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods. Futures in Mechanics of Structures and Materials: Proceedings of the 20th Australasian Conference on the Mechanics of Structures and Materials, ACMSM20. editor / Thiru Aravinthan ; Warna Karunasena ; Hao Wang. CRC Press, 2008. pp. 123-129
@inbook{d7299654aeb44f37988140911bddabf1,
title = "Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods",
abstract = "Developed for analysing long, periodic records, time series analysis methods are inherently suited for Structural Health Monitoring (SHM) applications. In this research, Autoregressive (AR) models were used to fit acceleration time histories from a 3-storey bookshelf laboratory structure and the ASCE Phase II SHM Benchmark Structure in healthy and several damaged states. Preliminary visual inspection of the large sets of AR coefficients to check the presence of clusters corresponding to different damage severities was achieved using Sammon mapping. Systematic classification of damage into states based on the analysis of the AR coefficients was achieved using two supervised classification techniques: Nearest Neighbour (NN) and Learning Vector Quantisation (LVQ), and one unsupervised technique: Self-Organising Maps (SOM).",
author = "{De Lautour}, {O. R.} and P. Omenzetter",
year = "2008",
month = "11",
day = "20",
doi = "10.13140/2.1.3574.4329",
language = "English",
isbn = "9780415491969",
pages = "123--129",
editor = "Thiru Aravinthan and Warna Karunasena and Hao Wang",
booktitle = "Futures in Mechanics of Structures and Materials",
publisher = "CRC Press",

}

TY - CHAP

T1 - Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods

AU - De Lautour, O. R.

AU - Omenzetter, P.

PY - 2008/11/20

Y1 - 2008/11/20

N2 - Developed for analysing long, periodic records, time series analysis methods are inherently suited for Structural Health Monitoring (SHM) applications. In this research, Autoregressive (AR) models were used to fit acceleration time histories from a 3-storey bookshelf laboratory structure and the ASCE Phase II SHM Benchmark Structure in healthy and several damaged states. Preliminary visual inspection of the large sets of AR coefficients to check the presence of clusters corresponding to different damage severities was achieved using Sammon mapping. Systematic classification of damage into states based on the analysis of the AR coefficients was achieved using two supervised classification techniques: Nearest Neighbour (NN) and Learning Vector Quantisation (LVQ), and one unsupervised technique: Self-Organising Maps (SOM).

AB - Developed for analysing long, periodic records, time series analysis methods are inherently suited for Structural Health Monitoring (SHM) applications. In this research, Autoregressive (AR) models were used to fit acceleration time histories from a 3-storey bookshelf laboratory structure and the ASCE Phase II SHM Benchmark Structure in healthy and several damaged states. Preliminary visual inspection of the large sets of AR coefficients to check the presence of clusters corresponding to different damage severities was achieved using Sammon mapping. Systematic classification of damage into states based on the analysis of the AR coefficients was achieved using two supervised classification techniques: Nearest Neighbour (NN) and Learning Vector Quantisation (LVQ), and one unsupervised technique: Self-Organising Maps (SOM).

UR - http://www.scopus.com/inward/record.url?scp=79952222744&partnerID=8YFLogxK

U2 - 10.13140/2.1.3574.4329

DO - 10.13140/2.1.3574.4329

M3 - Chapter (peer-reviewed)

SN - 9780415491969

SP - 123

EP - 129

BT - Futures in Mechanics of Structures and Materials

A2 - Aravinthan, Thiru

A2 - Karunasena, Warna

A2 - Wang, Hao

PB - CRC Press

ER -