Welding Pool Adaptive Vision Detection for Arc Welding Robot Gas Metal Arc Welding

YUE Jianfeng  LI Liangyu  FAN Fanglei  WANG Tianqi

(Tianjin Key Laboratory of Modern Mechatronics Equipment Technology, Tianjin Polytechnic University, Tianjin 300160)

Abstract: The research on image detection of welding pool for arc welding robot gas metal arc welding (GMAW) is made by a novel complementary metal oxide semiconductor (CMOS) vision sensor. Adaptive welding pool image collection is fulfilled by CMOS vision sensor’s circuit reconstructed. The high speed detection moment on welding pool image can be controllable after trigger circuit is reconstructed. The most suitable detection moment is found qualitatively by comparing welding pool images gained at different times. Exposure quantity of welding pool can be controllable after control circuit is reconstructed. The quantitative relation between welding feed speed (WFS) and exposure quantity is studied in the mode of WFS synergic control during GMAW. Condensation of welding pool gray value can be controllable after hard circuit is altered. The welding pool images collected during GMA-pulse welding and GMA short circuit welding are compared when CMOS vision sensor is working at different modes, and CMOS vision sensor working at linearity-logarithm mode is more suitable for GMAW pool image collection. Finally vivid welding pool image is both attained by integration application of above three technology to GMA pulse welding and GMA short-circuit welding separately, which make later image processing and welding quality controlling feasible.

Key words: Arc welding robot  Vision sensor  Welding pool image

CLC No: TG444+.73

天津市自然科学基金重点资助项目(05YFJZJ02100). Received 20070703, 20080118

References

[1] WU Lin, CHEN Shanben. Intellectualized Welding Technology[M]. Beijing: National Defence Industry Press, 2000.
[2] LUO H, DEVANATHAN R. Vision based neurofuzzy logic control of weld pool geometry[J]. Science and Technology of Welding and Joining, 2002, 7(5): 321-325.
[3] SUGA Y, MUKAI M, USUI S. Measurement of molten pool shape and penetration control applying neural network in TIG weld of thin steel plate[J]. ISIJ Internat, 1999, 39(10): 1 075-1 080.
[4] CHEN Shanben, CHEN Wenjie, LIN Tao. Identification models of weld pool dynamic process in pulsed GTAW[J]. Transactions of the China Welding Institution, 2001, 6(3): 5-8.
[5] CHEN Shanben, ZHAO Dongbin, LOU Yajun, et al. Intelligent control of weld seam molding in pulsed GTAW[J]. Transactions of the China Welding Institution, 2003, 6(1): 130-134.
[6] YAN Zhihong, ZHANG Guangjun, GAO Hongming, et al. Visual sensing of P-GMAW penetration information from multiple view angles[J]. Transactions of the China Welding Institution, 2006, 2(8): 47-50.
[7] WANG Qingyou. Image sensor application technology [M]. Beijing: Electron Industrial Publishing Company, 2006.