This research reviews the development and application of time domain and frequency domain transformations in the field of structural identification and particularly in the structural health monitoring (SHM) practices in the last half century. The challenges in conventional transformations and future trends are discussed in the light of the development of Time-Frequency domain techniques for the SHM of civil engineering structures. Fourier Transform, short time Fourier transform, Laplace transform, Wavelet transforms, and Hilbert transforms are exemplified as some currently used transformations in data mining studies in engineering practice. This presentation also reviews the basic principles of the transformation methods, which open new spectral view in understanding and feature extraction methodologies. The SHM operation covers the operational evaluation during monitoring the structure, data acquisition, fusion, cleansing, feature extraction, and decomposition. Time-Frequency approaches show that structural parameters are changing as observed from field measurements, values of the parameters are varying through the time recorded on real structures under constructed conditions with natural environment. Using Time-Frequency analysis techniques, system identification on a real structure may give the possibility to trace the signals to identify the structural damage and progressive deterioration including the detection of damage source and partially collapse mechanism. If instrumentation is dense enough, in local scale, in case of over member capacity exceedance, it may become possible that the priory condition before the crack or flaw initiation is detected at the available earliest stage. Based on the author's experience in the field, examples of results obtained from studies conducted over the years will be presented in this study.
Structural health monitoring (SHM),