First steps towards an intelligent laser welding architecture using deep neural networks and reinforcement learning

被引:85
作者
Guenther, Johannes [1 ]
Pilarski, Patrick M. [2 ]
Helfrich, Gerhard [1 ]
Shen, Hao [1 ]
Diepold, Klaus [1 ]
机构
[1] Tech Univ Munich, D-80333 Munich, Germany
[2] Univ Alberta, Dept Comp Sci, Edmonton, AB T6G 2E8, Canada
来源
2ND INTERNATIONAL CONFERENCE ON SYSTEM-INTEGRATED INTELLIGENCE: CHALLENGES FOR PRODUCT AND PRODUCTION ENGINEERING | 2014年 / 15卷
关键词
Deep learning; reinforcement learning; prediction; control; laser welding; REAL-TIME; BEHAVIOR;
D O I
10.1016/j.protcy.2014.09.007
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
To address control difficulties in laser welding, we propose the idea of a self-learning and self-improving laser welding system that combines three modern machine learning techniques. We first show the ability of a deep neural network to extract meaningful, low-dimensional features from high-dimensional laser-welding camera data. These features are then used by a temporal-difference learning algorithm to predict and anticipate important aspects of the system's sensor data. The third part of our proposed architecture suggests using these features and predictions to learn to deliver situation-appropriate welding power; preliminary control results are demonstrated using a laser-welding simulator. The intelligent laser-welding architecture introduced in this work has the capacity to improve its performance without further human assistance and therefore addresses key requirements of modern industry. To our knowledge, it is the first demonstrated combination of deep learning and Nexting with general value functions and also the first usage of deep learning for laser welding specifically and production engineering in general. This work also provides a unique example of how predictions can be explicitly learned using reinforcement learning to support laser welding. We believe that it would be straightforward to adapt our approach to other production engineering applications. (C) 2014 The Authors. Published by Elsevier Ltd.
引用
收藏
页码:474 / 483
页数:10
相关论文
共 28 条
[1]  
Alippi C, 2001, IEEE IMTC P, P1762, DOI 10.1109/IMTC.2001.929503
[2]   Optical sensor for real-time monitoring of CO2 laser welding process [J].
Ancona, A ;
Spagnolo, V ;
Lugarà, PM ;
Ferrara, M .
APPLIED OPTICS, 2001, 40 (33) :6019-6025
[3]  
[Anonymous], P 2010 INT C MECH AU
[4]  
[Anonymous], NIPS WORKSH DEEP LEA
[5]  
[Anonymous], ICALEO
[6]  
Bengio Yoshua, 2012, Neural Networks: Tricks of the Trade. Second Edition: LNCS 7700, P437, DOI 10.1007/978-3-642-35289-8_26
[7]   Whatever next? Predictive brains, situated agents, and the future of cognitive science [J].
Clark, Andy .
BEHAVIORAL AND BRAIN SCIENCES, 2013, 36 (03) :181-204
[8]  
Degris T, 2012, P AMER CONTR CONF, P2177
[9]  
Haykin S, 2012, P IEEE, V100, P3156, DOI 10.1109/JPROC.2012.2215773
[10]   Reducing the dimensionality of data with neural networks [J].
Hinton, G. E. ;
Salakhutdinov, R. R. .
SCIENCE, 2006, 313 (5786) :504-507