DeepWelding: A Deep Learning Enhanced Approach to GTAW Using Multisource Sensing Images

被引:78
作者
Feng, Yunhe [1 ]
Chen, Zongyao [1 ]
Wang, Dali [2 ]
Chen, Jian [3 ]
Feng, Zhili [3 ]
机构
[1] Univ Tennessee, Knoxville, TN 37996 USA
[2] Oak Ridge Natl Lab, Energy & Environm Sci Directorate, Oak Ridge, TN 37831 USA
[3] Oak Ridge Natl Lab, Phys Sci Directorate, Oak Ridge, TN 37831 USA
关键词
Welding; Neural networks; Deep learning; Monitoring; Feature extraction; Robot sensing systems; Arc welding; deep neural networks; gas tungsten arc welding (GTAW); monitoring and classification; multisource; pix2pix; sensing images; WELD JOINT PENETRATION; POOL SURFACE; PREDICTION; DEPTH;
D O I
10.1109/TII.2019.2937563
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Deep learning has great potential to reshape manufacturing industries. In this article, we present DeepWelding, a novel framework that applies deep learning techniques to improve gas tungsten arc welding process monitoring and penetration detection using multisource sensing images. The framework is capable of analyzing multiple types of optical sensing images synchronously and consists of three deep learning enhanced consecutive phases: image preprocessing, image selection, and weld penetration classification. Specifically, we adopted generative adversarial networks (pix2pix) for image denoising and classic convolutional neural networks (AlexNet) for image selection. Both pix2pix and AlexNet delivered satisfactory performance. However, five individual neural networks with heterogeneous architectures demonstrated inconsistent generalization capabilities in the classification phase when holding out multisource images generated with specific experimental settings. Therefore, two ensemble methods combining multiple neural networks are designed to improve the model performance on unseen data collected from different experimental settings. We have also found that the quality of model prediction is heavily influenced by the data stream collection environment. We think these findings are beneficial for the broad intelligent welding community.
引用
收藏
页码:465 / 474
页数:10
相关论文
共 35 条
[1]   ARTIFICIAL NEURAL NETWORKS APPLIED TO ARC-WELDING PROCESS MODELING AND CONTROL [J].
ANDERSEN, K ;
COOK, GE ;
KARSAI, G ;
RAMASWAMY, K .
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, 1990, 26 (05) :824-830
[2]  
[Anonymous], 2018, INT C LEARNING REPRE
[3]  
[Anonymous], 2008, Int. J. Adv. Manuf. Technol.
[4]   Prediction of gas metal arc welding parameters based oil artificial neural networks [J].
Ates, Hakan .
MATERIALS & DESIGN, 2007, 28 (07) :2015-2023
[5]  
Aytar Y, 2016, ADV NEUR IN, V29
[6]   A survey on intelligentized technologies for visual information acquirement, modeling and control of arc welding pool dynamics [J].
Chen, S. B. ;
Wu, J. .
IECON 2007: 33RD ANNUAL CONFERENCE OF THE IEEE INDUSTRIAL ELECTRONICS SOCIETY, VOLS 1-3, CONFERENCE PROCEEDINGS, 2007, :781-785
[7]  
Chen Shan-Ben., 2009, Intelligentized methodology for arc welding dynamical processes
[8]  
Chen Z, 2017, WELD J, V96, p367S
[9]   Welding penetration prediction with passive vision system [J].
Chen, Zongyao ;
Chen, Jian ;
Feng, Zhili .
JOURNAL OF MANUFACTURING PROCESSES, 2018, 36 :224-230
[10]   Intelligent Weld Manufacturing: Role of Integrated Computational Welding Engineering [J].
David, S. A. ;
Chen, Jian ;
Gibson, Brian T. ;
Feng, Zhili .
TRANSACTIONS ON INTELLIGENT WELDING MANUFACTURING, VOL I, NO. 2 2017, 2018, 1 (02) :3-30