Research And Application Of Welding Robot Recognition And Trajectory Planning

Manufacturing industry is the pillar industry of our country,welding is an important technology in manufacturing field.With the rapid development of China's industrial manufacturing industry,the requirements for the welding of workpieces are getting higher and higher,so the application of robot welding technology in manufacturing industry is becoming more and more extensive.At present,due to the lack of feedback characteristics of the mainstream demonstration reproducing welding robot,there will be welding trajectory deviation in welding parts,which will affect the quality of the product.At the same time,there are too many points to teach in the teaching process,which is very time-consuming,flexible and not strong.Aiming at the above problems,this paper studies welding robot weld identification and trajectory planning,which can help to improve the robot welding accuracy and automation level.Firstly,the paper introduces the experimental platform and then introduces the camera calibration.The calibration results are obtained by using the calibration toolbox in MATLAB,and the precision of binocular stereo vision measurement system is analyzed by binocular ranging experiment.In this paper,the image preprocessing technique is used,and a series of algorithms such as filter denoising,threshold processing and edge detection are combined to deal with the weld image.By comparing and analyzing the processing results of each operator,it is proved that Canny edge detection operator is suitable for edge detection operator in this paper.In this paper,the extraction technology of weld center line is studied.By analyzing the merits and demerits of the weld centerline extraction methods,it is concluded that the refined method is more suitable for the paper weld image centerline extraction method,and through this method to obtain the clear and continuous weld centerline.In this way,it can obtain the characteristic point coordinates of the center line of the weld image to guide the welding robot to complete the welding.Finally,the method of calculating the pixel coordinates of the feature points of the weld images is presented,and the three-dimensional space coordinates of the welding points are obtained by combining the calibration information and the welding torch position information.Then the trajectory planning of welding robot is realized.