Damage detection using enhanced multivariate statistical process control technique

Abstract

This paper addresses the problem of damage detection technique of structural health monitoring (SHM). Kernel principal components analysis (KPCA)-based generalized likelihood ratio (GLR) technique is developed to enhance the damage detection of SHM processes. The data are collected from the complex three degree of freedom spring-mass-dashpot system in order to calculate the KPCA model. The developed KPCA-based GLR is the method that attempts to combine the advantages of GLR statistic in the cases where process models are not available and a multivariate statistical process control

KPCA. The simulations show the improved performance of the KPCA-based GLR damage detection method.