Application of time series analysis and statistical pattern recognition for seismic damage detection

Piotr Omenzetter, Oliver R De Lautour

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

Abstract

Time series methods are inherently suited to the analysis of regularly sampled Structural Health Monitoring (SHM) data and deserve to be better and more extensively explored. This study focuses on the use of statistical pattern recognition techniques to classify seismic damage based on analysis of the time series model coefficients. Autoregressive (AR) models were used to analyze time histories from a 3-storey laboratory bookshelf structure excited on a shake table and the ASCE Phase II Experimental SHM Benchmark Structure in both healthy and damaged states. The coefficients of these AR models were used as damage sensitive features. Three supervised pattern recognition techniques, Back-propagation Artificial Neural Networks, Nearest Neighbor and Learning Vector Quantization were used to classify damage into states, quantify its severity and determine location. In order to visualize the data and reduce its dimensionality it was compressed using Principal Component Analysis or Sammon mapping. The minimum numbers of sensors required for reliable damage detection were also addressed. The results show that seismic damage can be detected and quantified by the three pattern recognition techniques with a very good accuracy using compressed data and small number of sensors.
Original languageEnglish
Title of host publicationProceedings of the 14th World Conference on Earthquake Engineering
Number of pages8
DOIs
Publication statusPublished - Oct 2008

Fingerprint

Time series analysis
Damage detection
Pattern recognition
Structural health monitoring
Time series
Vector quantization
Sensors
Backpropagation
Principal component analysis
Neural networks

Cite this

Omenzetter, P., & De Lautour, O. R. (2008). Application of time series analysis and statistical pattern recognition for seismic damage detection. In Proceedings of the 14th World Conference on Earthquake Engineering https://doi.org/10.13140/2.1.4360.8646

Application of time series analysis and statistical pattern recognition for seismic damage detection. / Omenzetter, Piotr; De Lautour, Oliver R.

Proceedings of the 14th World Conference on Earthquake Engineering. 2008.

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

Omenzetter, Piotr ; De Lautour, Oliver R. / Application of time series analysis and statistical pattern recognition for seismic damage detection. Proceedings of the 14th World Conference on Earthquake Engineering. 2008.
@inbook{ed72d852bbfe4d48934405ae57a69aba,
title = "Application of time series analysis and statistical pattern recognition for seismic damage detection",
abstract = "Time series methods are inherently suited to the analysis of regularly sampled Structural Health Monitoring (SHM) data and deserve to be better and more extensively explored. This study focuses on the use of statistical pattern recognition techniques to classify seismic damage based on analysis of the time series model coefficients. Autoregressive (AR) models were used to analyze time histories from a 3-storey laboratory bookshelf structure excited on a shake table and the ASCE Phase II Experimental SHM Benchmark Structure in both healthy and damaged states. The coefficients of these AR models were used as damage sensitive features. Three supervised pattern recognition techniques, Back-propagation Artificial Neural Networks, Nearest Neighbor and Learning Vector Quantization were used to classify damage into states, quantify its severity and determine location. In order to visualize the data and reduce its dimensionality it was compressed using Principal Component Analysis or Sammon mapping. The minimum numbers of sensors required for reliable damage detection were also addressed. The results show that seismic damage can be detected and quantified by the three pattern recognition techniques with a very good accuracy using compressed data and small number of sensors.",
author = "Piotr Omenzetter and {De Lautour}, {Oliver R}",
year = "2008",
month = "10",
doi = "10.13140/2.1.4360.8646",
language = "English",
booktitle = "Proceedings of the 14th World Conference on Earthquake Engineering",

}

TY - CHAP

T1 - Application of time series analysis and statistical pattern recognition for seismic damage detection

AU - Omenzetter, Piotr

AU - De Lautour, Oliver R

PY - 2008/10

Y1 - 2008/10

N2 - Time series methods are inherently suited to the analysis of regularly sampled Structural Health Monitoring (SHM) data and deserve to be better and more extensively explored. This study focuses on the use of statistical pattern recognition techniques to classify seismic damage based on analysis of the time series model coefficients. Autoregressive (AR) models were used to analyze time histories from a 3-storey laboratory bookshelf structure excited on a shake table and the ASCE Phase II Experimental SHM Benchmark Structure in both healthy and damaged states. The coefficients of these AR models were used as damage sensitive features. Three supervised pattern recognition techniques, Back-propagation Artificial Neural Networks, Nearest Neighbor and Learning Vector Quantization were used to classify damage into states, quantify its severity and determine location. In order to visualize the data and reduce its dimensionality it was compressed using Principal Component Analysis or Sammon mapping. The minimum numbers of sensors required for reliable damage detection were also addressed. The results show that seismic damage can be detected and quantified by the three pattern recognition techniques with a very good accuracy using compressed data and small number of sensors.

AB - Time series methods are inherently suited to the analysis of regularly sampled Structural Health Monitoring (SHM) data and deserve to be better and more extensively explored. This study focuses on the use of statistical pattern recognition techniques to classify seismic damage based on analysis of the time series model coefficients. Autoregressive (AR) models were used to analyze time histories from a 3-storey laboratory bookshelf structure excited on a shake table and the ASCE Phase II Experimental SHM Benchmark Structure in both healthy and damaged states. The coefficients of these AR models were used as damage sensitive features. Three supervised pattern recognition techniques, Back-propagation Artificial Neural Networks, Nearest Neighbor and Learning Vector Quantization were used to classify damage into states, quantify its severity and determine location. In order to visualize the data and reduce its dimensionality it was compressed using Principal Component Analysis or Sammon mapping. The minimum numbers of sensors required for reliable damage detection were also addressed. The results show that seismic damage can be detected and quantified by the three pattern recognition techniques with a very good accuracy using compressed data and small number of sensors.

U2 - 10.13140/2.1.4360.8646

DO - 10.13140/2.1.4360.8646

M3 - Chapter (peer-reviewed)

BT - Proceedings of the 14th World Conference on Earthquake Engineering

ER -