视觉传感螺旋钢管埋弧焊内焊焊缝自动跟踪系统

针对螺旋钢管埋弧焊内焊，研制一种基于视觉传感的焊缝自动跟踪系统。以视频摄像机为传感器，在焊接点前方300 mm处检测递送边钝边棱边。对视频信号进行放大、整形和二值化处理，获得对应于钝边棱边的脉冲信号，以帧中心为基准，利用单片机进行偏离方向判断及偏差值计算，并将偏差信号以脉冲方式送入交流伺服电动机驱动器，由交流伺服电动机控制焊枪滑移台移动，确保焊丝实时对准焊缝。试验结果表明：采用视频摄像机识别钝边棱边实现焊缝自动跟踪的方法是可行的。系统响应周期小于50 ms，跟踪精度达±0.3 mm。

Visual Sensor for Automatic Weld Seam Tracking System to Submerged Arc Inner Welding of Spirally Steel Tubes

Aimed at submerged arc inner welding of spirally steel tubes, a visual sensor for automatic weld seam tracking system is developed. Visual camera is used as sensor, the image of the side obtuse edge is taken at three hundred millimeters front of welded point. Video signal is magnified, trimmed and processed into binary, the pulse waveform of obtuse edge is obtained. Based on vertical center in video signals, the deviation from direction is judged and the deviation value is estimated by computer on-slice, and the deviation signals in the form of pulse signals are transmitted to the drive of AC servo motor, the motor will control the movement of torch sliding platform, so as to ensure the welding wire to be directed at the weld in real time. Experiment results show that the method is feasible to identify obtuse edge with visual camera; a reliable system for real-time seam tracking to submerged arc inner welding of spirally steel tubes is developed. The response cycle of this system is less than 50 ms, and the tracking precision is ± 0.3 mm.