Automatic damage classification based on wave cluster and principal component analysis
Document typeConference lecture
Rights accessOpen Access
Principal Component Analysis (PCA) plays a significant role in SHM field. There are plenty of algorithms that use PCA either directly or indirectly to detect damages in structures. Although PCA shows a successful role in damage detection but it still needs a complimentary step for automatic damage classification. It means a human effort still is required to classify different clusters that exists. Among different clas- sifiers, the wavelet classifier posses many dedicated merits. This work concentrates on automatic classification of damages with different severities. To do this, PCA is used as a tool for dimensionality reduction and then a wavelet classifier is applied on the result to classify different patterns in the structure each of which associated to significant state of the structure. This work involves experiments with composite plates powered by piezoelectric transducers as sensors and actuators. Damages are introduced into the structure as mass with different weights.
CitationGaribnezhad, F. [et al.]. Automatic damage classification based on wave cluster and principal component analysis. A: International Workshop on Structural Health Monitoring. "Structural Health Monitoring 2013: A Roadmap to Intelligent Structures". Stanford: 2013, p. 2760-2767.