Multi-pass Intelligent Welding Based On Robotic Control System Of Laser Vision Sensor

In the intelligent industrial production,welding robot has been widely used in the field of multi-pass automatic welding.The traditional robot is difficult for adapting to the change of the weld position and the groove size due to the influence of the workpiece positioning,the machining accuracy and the weld reinforcement.Therefore,it is a prerequisite to ensure the quality of multi-pass automatic welding that planning the torch trajectory according to the weld position and adjusting the welding process parameters according to the weld space size on the premise of teaching.In this paper,a robotic intelligent welding system based on laser vision sensor technology is designed by taking the multi-pass welding of Unilateral V-groove between low carbon steel shaft and flange as an example.The main work includes: Analyze the working principle of the sensor.A laser vision sensor is designed to detect the position of weld,measure the width and depth of groove.Determine the sensor parameters and detection accuracy.By analyzing the results of image processing methods and edge detectors,the processing algorithm is designed to realize the characteristic curve and the feature points.Plan the torch trajectory and design three welding methods including back,filler swing and cap welding.Dicede the basic process parameters.According to the change of weld bead space size,different swing process parameters are designed for filler welding,and different wire feeding speeds are designed for cap welding.The I/O communication between sensor system and robotic system is realized and the robotic welding program is designed.The robot modify the welding torch trajectory according to the seam position information,adjust the welding parameters during the filler welding process and adjust the wire feeding speeds during the cap welding process according to the weld reinforcement and groove width.The results show that the distance between the sensor and the workpiece is 60 ? 85 mm,the resolution is 0.0568 mm/pixel,the field of vision is 28.4 mm × 21.3 mm,the detection accuracy is 0.126 mm and the system response time is within 60 ms.By comparing the actual deviation with the correction results and measuring the weld cross-section size,the correction error of the system is less than 0.4 mm.By observing the actual welding process parameters and analyzing the weld appearance,it is concluded that the intelligent arc welding system can make the "teaching and palyback" robot obtain high quality weld seam in the multi-pass welding of shaft and flange unilateral V-groove.