Damage classification in structural systems using time series analysis and supervised and unsupervised clustering methods

Developed for analysing long, periodic records, time series analysis methods are inherently suited for Structural Health Monitoring (SHM) applications. In this research, Autoregressive (AR) models were used to fit acceleration time histories from a 3-storey bookshelf laboratory structure and the ASCE Phase II SHM Benchmark Structure in healthy and several damaged states. Preliminary visual inspection of the large sets of AR coefficients to check the presence of clusters corresponding to different damage severities was achieved using Sammon mapping. Systematic classification of damage into states based on the analysis of the AR coefficients was achieved using two supervised classification techniques: Nearest Neighbour (NN) and Learning Vector Quantisation (LVQ), and one unsupervised technique: Self-Organising Maps (SOM).