Acta Photonica Sinica, Volume. 50, Issue 12, 1212004(2021)
Debonding Defect Identification Method for Multi-layer Bonded Structures Based on LDA-CPSO-SVM Optimization
Figures & Tables(9)
Fig. 1. Design drawings of multilayer bonded structure samplesⅠand Ⅱ
Fig. 3. Terahertz time domain waveform of different regions of multilayer adhesive structure
Fig. 5. Classification rate fitness curve of adhesive layer I data
Fig. 7. Terahertz single feature image of multi-bonded structural adhesive layer
Fig. 8. Imaging image of multi-bonded structural adhesive layer based on LDA-CPSO-SVM algorithm
Table 1. Classification rate of different kernel function types
View tableView in ArticleTable 1. Classification rate of different kernel function types
Sample type The accuracy of different kernel function types Liner SVM Quadratic SVM Cubic SVM Fine Gaussian SVM Medium Gaussian SVM Coarse Gaussian SVM Layer I normal 85.22% 86.51% 84.34% 88.95% 96.13% 94.66% Layer I defect 86.45% 87.36% 89.65% 88.39% 99.43% 98.9% Layer I edge 85.85% 86.37% 89.32% 89.67% 98.34% 97.67% LayerⅡnormal 86.27% 87.57% 90.34% 91.98% 95.14% 97.66% LayerⅡdefect 83.42% 89.35% 88.66% 90.38% 98.49% 98.96% LayerⅡedge 81.84% 86.2% 88.33% 91.66% 96.3% 97.37%
Tools
Get Citation
Copy Citation Text
Li ZHENG, Chuang LIU, Jiaojiao REN, Dandan ZHANG, Lijuan LI, Jisheng XU. Debonding Defect Identification Method for Multi-layer Bonded Structures Based on LDA-CPSO-SVM Optimization[J]. Acta Photonica Sinica, 2021, 50(12): 1212004
Paper Information
Category: Instrumentation, Measurement and Metrology
Received: Jun. 25, 2021
Accepted: Aug. 25, 2021
Published Online: Jan. 25, 2022
The Author Email: LIU Chuang (liuchuang@cust.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20