A New Inverse Method of Elastic Constants for a Fibre-Reinforced Composite Plate from Laser-Based Ultrasonic Lamb Waves

A new inverse method based on the wavelet transform and artificial neural networks (ANN) is presented to recover elastic constants of a fibre-reinforced composite plate from laser-based ultrasonic Lamb waves. The transient waveforms obtained by numerical simulations under different elastic constants are taken as the input of the ANN for training and learning. The wavelet transform is employed for extracting the eigenvectors from the raw Lamb wave signals so as to simplify the structure of the ANN. Then these eigenvectors are input to a multi-layer internally recurrent neural network with a back-propagation algorithm. Finally, the experimental waveforms are used as the input in the whole system to inverse elastic constants of the experimental material.