Enhanced features in principal component analysis with spatial and temporal windows for damage identification

Principal component analysis (PCA) methods have been widely applied to damage identification in the long-term structural health monitoring (SHM) of infrastructure. Usually, first few eigenvector components derived by PCA are treated as damage-sensitive features. In this paper, effective method double-window (DWPCA) and novel features proposed for better performance. method, spatial temporal windows introduced traditional method. The group sensors exclude those insensitive damage, while window is discriminate eigenvectors between damaged healthy states. addition, length directional angle variation states used features, instead previous studies. Numerical simulations based on a large-scale bridge reveal that successful identifying located far from due use both well variation. compared moving methods, higher sensitivity resolution identification.