Structural damage identification by using wavelet entropy

The wavelet transform is combined with Shannon entropy to detect structural damage from measured vibration signals in this paper. Damage features such as wavelet entropy, relative wavelet entropy and wavelet-time entropy are defined and investigated to detect and locate damage. The damage identification method is examined by a numerically simulated case and two laboratory test cases. It is demonstrated that wavelet-time entropy is a sensitive damage feature in detecting the abnormality in measured successive vibration signals, while relative wavelet entropy is a good damage feature to detect damage occurrence and damage location through the vibration signals measured from the intact (reference) and damaged structures. In addition, the relative wavelet entropy method is flexible in choosing the reference signal simultaneously measured from any undamaged location of the target structure. This advantage is particularly interesting in detecting the damage of existing structures because data from the intact (undamaged) structure is not required.