BILINEAR TIME-FREQUENCY TRANSFORMATIONS IN THE ANALYSIS OF DAMAGED STRUCTURES

The performances of six time-frequency transforms applied to the asses sment of structural damage are compared and their properties discussed from a theoretical point of view. Some examples of applications to sy nthetic and real signals are presented. Specifically, their diagnostic sensitivity when processing accelerometer signals recorded during the impulsive loading of damaged aluminum beams is discussed. The test pr ogramme involved different position and damage levels artificially pro duced on the beams. The results obtained outline the importance of fin ding criteria for a fully automatic interpretation of the time-frequen cy transforms. Also the problems associated with the use of neural pat tern recognition in the diagnostic interpretation of time-frequency ma ps of signals detected on structures are illustrated. An application s hows the results obtained in the neural classification of structural d efects by applying techniques based on invariant moments. (C) 1997 Aca demic Press Limited.