Seam Tracking Control Based On DSC

Because of the intelligent, automation and flexibility technology developing fast, welding has experienced a significant leap from manual to automated and intelligent welding. But at the same time, on the welding itself, the welding costs and efficiency requirements are also becoming more demanding. Therefore, the continuous improvement of welding automation and intelligent level, for the welding industry to reduce costs, improve efficiency and enhance the competitiveness of great significance. Under the intelligently welding target, the paper has the following aspects:Firstly, using SAF automatic welding carriage as motion platform, Oldham for the movement of the actuator, binocular CCD as the visual sensor, combined dsPIC30F5015 digital signal controller with PC as the lower and upper machine. Build a hardware and software of welding robot seam tracking control system. Taking advantage of MATLAB GUI prepared image processing and motion control system which can make calibration, image processing, weld fitting, error signal extraction and communications come true.Secondly, after calibrating the binocular CCD, extracting a lot of seam image, developed a kind of image processing method which suit for the present experiment environment. Taking advantage of multi-structural elements morphology to extract real weld edge, find the centerline, combined with Hough, extraction of the weld initial point.Finally, using the results of image processing, reconstruction the weld seam, get the world coordinate of centerline and feature points pixels. Based on dsPIC30F5015 and improved point by point comparison, interpolate the centerline, while the interpolation calculation process, PC extract the torch and the weld centerline deviation, combined with PID algorithm, generate correction signal. Implement seam tracking function.