Layered defect imaging method and system using chirp lamb waves and fully convolutional networks
By combining linear frequency modulated Lamb wave excitation with a fully convolutional neural network, the problem of the inability to comprehensively detect delamination defects in carbon fiber plates in existing technologies is solved, achieving accurate imaging and identification of defects of different sizes and reducing model training costs.
Patent Information
- Authority / Receiving Office
- CN Β· China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUAZHONG UNIV OF SCI & TECH
- Filing Date
- 2024-12-10
- Publication Date
- 2026-06-02
AI Technical Summary
Existing technologies cannot comprehensively detect delamination defects in carbon fiber plates of different sizes, and existing linear frequency modulated Lamb wave imaging methods cannot obtain accurate defect imaging images.
By employing linear frequency modulated Lamb wave excitation combined with a fully convolutional neural network, a Lamb wave signal is acquired through a piezoelectric transducer and subjected to wavelet transform to construct an imaging model. The fully convolutional neural network is then used to extract the time-frequency domain signal features of the Lamb wave, thereby achieving accurate imaging of defects.
It enables accurate imaging and identification of delamination defects in carbon fiber plates, reduces model training costs, and improves the accuracy and comprehensiveness of detection.
Smart Images
Figure CN119648835B_ABST