Research On Laser Vision Sensor Of Seam Detection And Automatic Seam Tracking System

In welding process, there mus be offset between welding torch and seam center. As everyone knows, because of welding process is a complicated, nonlinear process with multi-interference, thus thermal deformation, unfitness of butt joint, undercut, and the change of seam gap length, which may influence welding quality directly, are all unpredictable.Therefore, real time seam tracking can improve welding precision, especially suitable for the welding of products, which can lead to thermal deformation easily, like aluminium alloy sheets which been used in this article.The essay based on project of provincial science and technology commission, with a background of weld of mild steel and aluminium alloy, aimed to the method of MIG, and did weld tracking experiment on straight butt joint, lap joint, corner joint of mild steel and aluminium alloy sheets. In order to achieve seam tracking control, the seam detection is one of most important technologies. The article analyzes the noise types and the reasons which can lead to affect welding images process. We have introduced the methods of image process, and compared the methods which recommended in article. Through compared the welding image process and the results of experiments, in order to improve speed of welding images process and the precision of weld feature points, we have came up with a more effective and accurate solution for image process.At the same time, we analyzed welding images which collected in the welding process; the result shows that welding spatter and arc light are most factors that affected the image quality. To solve the problem, the article proposed that we equipped with an optical filter in the front of CCD camera which have specific wavelength.Meanwhile, the mask and the filtering algorithms are also essential.The project used an equipment of welding tracking system, which contained laser vision sensor, step drive and IPC control system, and the equipment can realize automatic detection and tracking control. In the case of open seam tracking or not, with the background of mild steel and aluminium alloy, the author used software of VC++6.0 image process to detect the welding torch coordinates. The result which through analysis and compare from experimental data shows that, the welding torch always rotates around seam center when the tracking system is used, and the data shows that the tracking precision is higher than ±0.69mm. It is worth mentioning that the appearance of weld is well through actual welding. In addition, the system can realize automatic detection of unfitness butt joint. The result shows that the detection precision and sensitivity of automatic system meet requirement for the practical production. The lase vision sensing seam tracking system is not only improve welding quality and welding efficiency, but also decrease intensity of labour and defective index. What’s more, it would contribute to the cost reduction and energy saving at the same time.