Acoustic Emission Testing Research of Composites Bearing Based on Neural Network

摘要

This paper will apply the Acoustic Emission(AE) technique principle to detect the AE signals of the three-dimensional braided composites under tension and compression test mode and apply wavelet analysis to reduce the AE signal noise. The filtered AE waveform or waveform parameters will be treated as a sample to be input to Back Propagation(BP) neural network, after the training, BP neural network will automatically identify the load bearing of three-dimensional braided composite materials and its corresponding damage model.