Published: 31 March 2012

Damage detection based on pattern-matching method of response signal

Jin Bong Kim1
Hee-Chang Eun2
Views 0
Reads 0
Downloads 1039

Abstract

Using response data as the feature to extract the damage, this study proposes an analytical method to detect damage by matching the response pattern between a damage-expected beam system and the database. Response pattern database is composed of the response variation before and after the damage, prescribed by various damage scenarios. Pattern-matching is evaluated by the correlation coefficients between two data sets and the damage corresponds to high correlation coefficient. Response variations at a full set of DOFs are estimated by the expansion of a few measured response data. Numerical applications illustrate the validity and applicability of the proposed method.

About this article

Received
21 November 2011
Accepted
14 February 2012
Published
31 March 2012
Keywords
pattern-matching
damage detection
feature vector
correlation
constraint force