EMD-based pulsed TIG welding process porosity defect detection and defect diagnosis using GA-SVM
详细信息 查看全文
- 作者：Yiming Huanga ; Di Wua ; Zhifen Zhangb ; Huabin Chena ; hbchen@sjtu.edu.cn" class="auth_mail" title="E-mail the corresponding author ; Shanben Chena ; sbchen@sjtu.edu.cn" class="auth_mail" title="E-mail the corresponding author
- 关键词：Arc spectrum ; EMD ; Feature extraction ; Hydrogen porosity ; GA-SVM
- 刊名：Journal of Materials Processing Technology
- 出版年：2017
- 出版时间：January 2017
- 年：2017
- 卷：239
- 期：Complete
- 页码：92-102
- 全文大小：3317 K
文摘
- •
A portable spectrometer that can acquire and process arc spectrum data in real-time is designed.
- •
An innovative method based on spectral analysis is proposed to perform porosity defect detection.
- •
The relationship between features extracted from spectral lines and porosity size is investigated.
- •
A GA-SVM model is built to classify different types of porosity.
- •
Experiments based on X-ray inspection are carried out to verify the proposed technique.