Vibration-based structural condition monitoring using concept of digital damage fingerprints (DDF)

The practicability of vibration-based structural condition and health monitoring is often challenged due to the difficulty in cost-effectively understanding the complicated vibration signals. Aimed at interrogating the vibration signals in an explicit manner, a signal processing and interpretation scheme capitalising on advanced signal processing and data fusion approaches was developed, termed digital damage fingerprints (DDF). In the approach, wavelet transform and artificial neural network algorithms were employed to sequentially conduct signal pre-processing, purification, characteristic extraction, data compression, information mapping and pattern recognition, whereby essential yet concise features were extracted from the acquired vibration signal, to further assist in the evaluation of structural condition and health status. For validation, the proposed approach was employed to quantitatively assess delamination in composite structures, and the results exhibited excellent capability of the approach in facilitating real-time surveillance on structural condition.