Classification of Plume Image and Analysis of Welding Stability during High Power Disc Laser Welding

Abstract:

Plume images which captured during the high power disc laser welding contain lots of information that related to the welding quality and stability. The classification of plume images is an important foundation for online monitoring during welding process. Stainless steel 304 was taken as the experiment object for the high power disc laser welding experiments. A high-speed camera was used to capture the ultraviolet band and visible light band plume images in the laser welding process. Image processing techniques were applied to get the characteristic parameters such as the ratio of the absolute value of coordinate difference between the centriod of plume and the welding point, the number of spatter, the average gray level and entropy of a spatter image, and formed a four dimension vector. Then K-nearest neighbor classification method was used to separate the poor welding quality images out from good ones. Welding experimental results confirmed that using K- nearest neighbor classification method to classify the four dimension vector samples which formed by the ratio of absolute value of coordinate difference between the centriod of plume and welding point, number of spatters, average gray level and entropy could obtain a recognition rate that close to the actual welding results.

• Periodical: Applied Mechanics and Materials (Volumes 201-202)
• Pages: 1139-1142
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.201-202.1139
• Online since: October 2012
• Authors: Qian Wen, Xiang Dong Gao
• Keywords: Classification, High Power Disk Laser Welding, Plume Image, Welding Stability

References

[1] X. D. Gao, L. Mo, X.G. Zhong, D.Y. You and S. Katayama: Acta Physica Sinca, (2011) No.8, pp.737-744.

[2] X.D. Gao , W.X. Lv and D.Y. You: Transactions of China Welding Institution, Vol.32 (2011), pp.5-8 (in Chinese).

[3] B. Ji: Computer Technology And Development, Vol.18(2008) No.6,pp.73-75 (in Chinese).

[4] X.G. Zhang: Pattern Recognition (Tsinghua University Press, China 2010).