Study On On-line Monitoring And Weld Surface Quality Inspection System For High-speed Vehicle Body Welding

Combined with the automation and intelligent requirements of high-speed car body welding process,this paper studied welding seam tracking technology,welding speed adjustment technology and weld surface quality detection technology based on 2D laser displacement sensor and machine vision.The above technologies were tested by field experiment and the application effect was analyzed and verified.The main research contents of this paper are as follows:(1)Welding seam real-time tracking technologyCombined with the requirements of weld seam tracking in car body welding production,a detection and control method combining 2D laser displacement sensor and machine vision was proposed.Firstly,the laser sensor on the front of the welding gun was used to detect and locate the center of the groove.Then the visual sensor was used to collect the information of the weld pool.The real-time offset of wire center and groove center was processed by the upper computer,and the results of laser feed-forward guidance were modified to improve the accuracy of welding seam tracking.(2)Real-time welding speed adjustment technologyA method for determining the welding speed from the groove area was proposed.The method detected the section information of the groove by the laser displacement sensor,and calculated the boundary curve of the groove section and the area enclosed by the reference line.The ratio of the area to the standard cross-sectional area was measured in real time through the groove,and the adjustment parameters of the welding speed were calculated to improve the welding quality.(3)Automatic inspection technology of weld surface qualityAccording to the requirement of automatic detection of weld surface quality,this paper presented a detection method based on laser vision.The method used laser sensor to obtain weld section information.The judging conditions were set according to the evaluation criteria and specifications of different weld surface defects.After processing,the upper computer determined the type of defect and gave the specific parameters.When the defect occurs,the upper computer read the pulse number of the welding robot's gantry motor through the motion control card and obtained the current position of the welding robot,and realized the localization of the defect.The research results of this paper have certain theoretical significance and practical value for enriching automatic welding technology as well as promoting the research of intelligent welding equipment based on machine vision.