Considerations for practical neural network application to a damage detection problem

S.G. Pierce, K. Worden, G. Manson

Research output: Contribution to journalArticle

Abstract

The application of a multilayer perceptron (MLP) neural network to a damage location problem on a GNAT aircraft wing is considered. The problems associated with effective network training and evaluation are discussed, focussing on ensuring good generalisation performance of the network to the classification of new data. Both conventional Maximum Likelihood and Bayesian Evidence based training techniques are considered and a simple thresholding technique is presented to aid in the rejection of poorly regularised network structures. Examples are presented for an artificial simple 2 class problem (drawn from a Gaussian distribution) and a real 9 class problem on the GNAT aircraft wing.
LanguageEnglish
Pages151-158
Number of pages7
JournalKey Engineering Materials
Volume293
DOIs
Publication statusPublished - 2005

Fingerprint

Damage detection
Neural networks
Gaussian distribution
Multilayer neural networks
Maximum likelihood

Keywords

  • damage detection
  • neural networks
  • regularisation
  • generalisation
  • thresholding
  • power systems

Cite this

@article{72e1011dcda246d887279d7e4142651a,
title = "Considerations for practical neural network application to a damage detection problem",
abstract = "The application of a multilayer perceptron (MLP) neural network to a damage location problem on a GNAT aircraft wing is considered. The problems associated with effective network training and evaluation are discussed, focussing on ensuring good generalisation performance of the network to the classification of new data. Both conventional Maximum Likelihood and Bayesian Evidence based training techniques are considered and a simple thresholding technique is presented to aid in the rejection of poorly regularised network structures. Examples are presented for an artificial simple 2 class problem (drawn from a Gaussian distribution) and a real 9 class problem on the GNAT aircraft wing.",
keywords = "damage detection, neural networks, regularisation, generalisation, thresholding, power systems",
author = "S.G. Pierce and K. Worden and G. Manson",
year = "2005",
doi = "10.4028/0-87849-976-8.151",
language = "English",
volume = "293",
pages = "151--158",
journal = "Key Engineering Materials",
issn = "1013-9826",

}

Considerations for practical neural network application to a damage detection problem. / Pierce, S.G.; Worden, K.; Manson, G.

In: Key Engineering Materials, Vol. 293, 2005, p. 151-158.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Considerations for practical neural network application to a damage detection problem

AU - Pierce, S.G.

AU - Worden, K.

AU - Manson, G.

PY - 2005

Y1 - 2005

N2 - The application of a multilayer perceptron (MLP) neural network to a damage location problem on a GNAT aircraft wing is considered. The problems associated with effective network training and evaluation are discussed, focussing on ensuring good generalisation performance of the network to the classification of new data. Both conventional Maximum Likelihood and Bayesian Evidence based training techniques are considered and a simple thresholding technique is presented to aid in the rejection of poorly regularised network structures. Examples are presented for an artificial simple 2 class problem (drawn from a Gaussian distribution) and a real 9 class problem on the GNAT aircraft wing.

AB - The application of a multilayer perceptron (MLP) neural network to a damage location problem on a GNAT aircraft wing is considered. The problems associated with effective network training and evaluation are discussed, focussing on ensuring good generalisation performance of the network to the classification of new data. Both conventional Maximum Likelihood and Bayesian Evidence based training techniques are considered and a simple thresholding technique is presented to aid in the rejection of poorly regularised network structures. Examples are presented for an artificial simple 2 class problem (drawn from a Gaussian distribution) and a real 9 class problem on the GNAT aircraft wing.

KW - damage detection

KW - neural networks

KW - regularisation

KW - generalisation

KW - thresholding

KW - power systems

UR - http://new.ttp.net/0-87849-976-8/151/

UR - http://dx.doi.org/10.4028/0-87849-976-8.151

U2 - 10.4028/0-87849-976-8.151

DO - 10.4028/0-87849-976-8.151

M3 - Article

VL - 293

SP - 151

EP - 158

JO - Key Engineering Materials

T2 - Key Engineering Materials

JF - Key Engineering Materials

SN - 1013-9826

ER -