Model-Based Neural Network and Wavelet Packets Decomposition on Damage Detecting of Composites
Model-based neural network (MBNN) is used along with wavelet packets decomposition to detect internal or hidden damage in composites. In consideration of internal delaminations with different sizes and locations typical finite element model is used to acquire training data of neural networks. Delamination-induced energy variations are decomposed by wavelet packets to enhance damage features. The predicted delamination size and location are selected as output of neural networks. In order to acquire target signals, forced vibration test is conducted. Based on experimental result, damage-induced energy variation of response signal is analyzed and the relationship between damage and physical performance is related. Test result shows that the proposed method is effective to investigate internal damage state in composites.
KeywordsMBNN composites delamination detecting wavelet packets
Unable to display preview. Download preview PDF.
- 1.Bishop, C.M.: Neural Networks for Pattern Recognition. Oxford University Press, New York (1998)Google Scholar
- 2.Cios, K.J., Tjia, R.E.: Application of Neural Networks in the Acousto-ultrasonic Evaluation of Metal-Matrix Composite Specimens. In: International Joint Conference on Neural Networks, Singapore, vol. 2, pp. 993–998 (1992)Google Scholar
- 5.Yang, J., Cai, N., Hu, K., Xiong, H.: MRF-MBNN: A Novel Neural Network Architecture for Image Processing. In: Wang, J., Liao, X.-F., Yi, Z. (eds.) ISNN 2005. LNCS, vol. 3497, pp. 673–678. Springer, Heidelberg (2005)Google Scholar