Research On Control Algorithms Used For Motion Control Of Welding Torch During Arc Welding

Arc welding is a very important manufacturing metal joining process, which is widely used in the mechanical manufacturing, aerospace, automotive, shipbuilding, bridges and nuclear industry and other industries. In recent years, with the improvement and demand of the quality of the welding weldment, changes in the field of welding are imperative. And the traditional manual welding will eventually be eliminated. Therefore, the automation and intelligence of welding process is the trend of development in the field of welding in the future.The technology of seam tracking is an important area in the research of automatic arc welding. Precise seam tracking is of great importance to ensure the quality of automatic acc welding. To achieve precise seam tracking, the detection of seam deviation (ie, the weld deviation of the center of the weld and the center of the arc) is a key prerequisite. In arc welding process of seam tracking, the motion control system of the welding torch could be the control object. Welding torch relative to the center position deviation is stressed by means of a variety of sensors such as mechanical, electromagnetic, laser, vision. And the the system controls the welding torch in the process so that it is always on the weld.Arc welding is essentially a complex thermal process. In the welding process, interference will be produced such as light, heat and thermal deformation of the workpiece so that it is difficult to achieve precise seam tracking. In the present automatic welding apparatus, the movement of the torch are controlled by a stepper motor and drive mechanism to be achieved, thus improving the precision of control system is the key technology to achieve accurate seam tracking.To ensure the quality of weldment, weld path must to be accurate tracking. People used a variety of techniques to get the center of the track welds. And the technology of weld identification and location which is based on machine vision have been widely adopted,because of its intuitive visual system and the amount of information collected. This paper is research on the use of tungsten inert gas welding of welding which is based on visual sensors. The system includes a stepper motor and drive, vision sensors and image acquisition card, table and computer and other devices. First, the paper analyzes the characteristics of the welding images and obtains the position information of the images of weld pool by image processing techniques so as to extract the position deviation of the weldments, providing the basis for the realization of seam tracking. Then, according to the motion characteristics and principles of the actuator device,the mathematical model for the torch motion control system is established. In order to establish a more accurate mathematical model, in the seam tracking system, the controlled object includes two-phase hybrid stepping motor, roller screw, moving table and weldments, fixtures and other loads. Finally, computer simulation by traditional PID combined with radial basis function neural network was to be finished by the soft of MATLAB.Then the tracking performance of the system is improved and lays a good foundation for the actual seam tracking tests.