Development Of Tracking Equipment And Experiment Of Technology Based On Laser Vision In High-Speed Welding

In this paper, with regard to the problem that the amount of deviation existed in seam tracking for high-speed welding is too large, a set of tracking equipment based on laser vision was developed, which achieved seam tracking for high-speed welding accurately.The most suitable structure of the vision sensor which based on structured light was proposed; the physical devices were selected and analyzed to build on the hardware structure of tracking equipment.The mounting fixture for CCD camera was designed and improved. The optimum design parameters of the vision sensor were obtained through the experiments:CCD installation height is 145mm, the installation height of the laser is 140mm, and the laser deflection angle is 20 degrees. The system was calibrated through different coordinates and the corresponding relationship between the pixel and the actual size of the image was established. It was confirmed that 6 pixels in the figure correspond to the actual size of 1mm through measuring.The laser seam tracking software was developed under VC++6.0 development platform. The main control parameters of the system include welding method, welding material, groove type, joint type and welding speed.In order to make the image processing both achieve accuracy and real-time performance, the image processing algorithm was respectively designed for the condition of no-arc and real time welding, made the entire process time in less than 50ms, which fully met the requirements of real-time seam tracking.Aimed at the shortcoming of common feature point searching method, this paper proposed maximum distance method which is suitable for all types of groove. The basis to distinguish various groove was proposed according to the feature point coordinates and the flow chart of internal identification was designed which achieved the automatic recognition of groove. According to the recognition criteria proposed in this paper, through the recognition test on various types of groove, it was found that the maximum distance method is very suitable for groove recognition and met the requirements of real-time tracking of welding.Finally, to verify the performance of tracking equipment based on laser vision developed in this paper, the tests of weld seam tracking under the condition of no-arc and image processing in MIG welding and real time welding tracking were carried out. At last, the results showed that when the welding speed was in 0.90m/min 1.14m/min and the slope of joints were in the 10°～20°, the weld seam tracking deviation was concentrated in 0.3 -0.6mm; when the welding speed was 0.90m/min, the deviation of curve seam was concentrated in 0.6mm and the maximum deviation was 0.8mm, which met the requirements of practical application.