Intelligent Damage Detection of Composite Structure Based on Convolutional Neural Network and Wavelet Transform

Abstract

Carbon fiber reinforced plastics (CFRP) composites has been widely used in aerospace, automotive and other fields due to its superior comprehensive performance. To guard against possible risks, nondestructive testing (NDT), such as Lamb wave has been used to monitor the internal damage of the CFRP composites structure and develop the dependable structural health monitoring techniques. However, since the high degree of anisotropy in the CFRP composites makes the failure mode under the fatigue load complicated, an accurate and timely damage detection of the CFRP composites is still difficult. This paper proposes a new damage detection method using convolutional neural network and continuous Wavelet transform. An experimental study is implemented to verify the effectiveness of the proposed method. The result shows that the proposed methodology is a useful technique for precise detection of delamination damage in the composites structure.