On-line learning of welding bead geometry in industrial robots
- 336 Downloads
In this paper, we propose an architecture based on an artificial neural network (ANN), to learn welding skills automatically in industrial robots. With the aid of an optic camera and a laser-based sensor, the bead geometry (width and height) is measured. We propose a real-time computer vision algorithm to extract training patterns in order to acquire knowledge to later predict specific geometries. The proposal is implemented and tested in an industrial KUKA KR16 robot and a GMAW type machine within a manufacturing cell. Several data analysis are described as well as off-line and on-line training, learning strategies, and testing experimentation. It is demonstrated during our experiments that, after learning the skill, the robot is able to produce the requested bead geometry even without any knowledge about the welding parameters such as arc voltage and current. We implemented an on-line learning test, where the whole experiments and learning process take only about 4 min. Using this knowledge later, we obtained up to 95 % accuracy in prediction.
KeywordsANN Computer vision Bead geometry Industrial robots Robotics on-line learning Robotics welding
Unable to display preview. Download preview PDF.
- 1.World Robotics 2013 (2013) Industrial Robots, International Federation of RoboticsGoogle Scholar
- 2.Gorle P, Clive A (2013) Positive impact of industrial robots on employment, international federation of robotics, METRA MARTECH LimitedGoogle Scholar
- 6.Cudina M, Prezelj J, Polajnar I (2008) Use of audible sound for on-line monitoring of gas metal arc welding process. Metalurgia 47(2):81–85Google Scholar
- 8.Kolahan F, Heidari M (2010) A New Approach for Predicting and Optimizing Weld Bead Geometry in GMAW. International. J Mech Syst Sci Eng 2:2:138–142Google Scholar
- 9.Sreeraj P, Kannan T, Maji S (2013) Prediction and control of weld bead geometry in gas metal arc welding process using simulated annealing algorithm. Int J Comput Eng Res 3(1):213–222Google Scholar
- 11.Grossberg S, Markuzon N, Reynolds JH, Carpenter GA, Rosen DB (1992) Fuzzy ARTMAP: a neural network architecture for incremental supervised learning of analog multidimensional maps. IEEE Trans Neural Netw 3(5)Google Scholar
- 14.Iqbal A, Khan SM, Mukhtar HS (2011) ANN assisted prediction of weld bead geometry in gas tungsten arc welding of HSLA steels. In: Proceedings of the World Congress on Engineering, Vol I, WCE 2011, July 6–8, 2011, London, U.KGoogle Scholar
- 16.Chan B, Pacey J, Bibby M (1999) Modelling gas metal arc weld geometry using artificial neural network technology. Can Metall Q 38(1):43–51Google Scholar
- 19.Sreeraj P, Kannan T (2012) Modelling and prediction of stainless steel clad bead geometry deposited by GMAW using regression and artificial neural network models. Adv Mech Eng 2012:1–12. Article ID 237379Google Scholar
- 20.Akkas N, Karayel D, Ozkan SS, Ogur A, Topal B (2013) Modeling and Analysis of the Weld Bead Geometry in Submerged Arc Welding by Using Adaptive Neurofuzzy Inference System Mathematical Problems in Engineering, Vol. 2013, Article ID 473495, 10 pagesGoogle Scholar