Research On Key Technology Of Corrugated Plate Welding Based On Visual Trajectory Recognition And Arc Tracking

The marine equipment represented by container corrugated board has problems such as low welding automation and poor welding quality.The effect of trajectory recognition and welding gun attitude recognition in practical applications is not satisfactory.The main reason is that the container is a large structural part.It is difficult to clamp,the corrugated board used for the side board has a large board surface,complex waveform changes,and the shape of the weld seam is difficult to maintain.The thickness of the board is generally 1 to 2 mm.It is easy to produce wave-shaped thermal deformation when heated.It can be up to 15 meters,and it is easy to bend under its own gravity.Therefore,the welding seam between corrugated plate and bottom side beam is a complex space welding seam,so that it is difficult to identify the weld track and the welding gun attitude.Aiming at the problems existing in the corrugated board trajectory recognition and welding seam tracking in the above large structural parts container,this paper proposes a corrugated board welding seam tracking method based on laser visual trajectory recognition and arc tracking.Use laser vision sensor to obtain the spatial point cloud data of the corrugated plate weld,extract the spatial corner data set of the weld from the point cloud data,and obtain the spatial weld seam trajectory,trajectory normal vector,and inflection point position by data processing of the corner point data And the angle of the broken line,etc,so as to predict the welding path of the corrugated board and adjust the welding posture.For large deviations caused by thermal deformation and mechanical wear during the welding process,an arc sensor is used to correct in real time to achieve corrugated board welding Prediction of seam trajectory and accurate weld seam tracking,specific research contents are as follows:1)Designed an experimental platform for corrugated plate welding seam tracking based on the method proposed in this paper.Modules such as PCI1020 Altai motion control card,PCI8735 Altai data acquisition card,LJ-G200 Keyence laser vision sensor,turbine worm turntable and oscillator are used to design the experiment platform for corrugated plate welding point cloud data acquisition,trajectory identification and welding seam tracking.2)Corrugated board weld trajectory recognition and welding gun attitude recognition based on point cloud data processing.Analyze different corrugated plates,such as trapezoidal welds,arc-shaped welds,and establish their corner and inflection points models.Using the HARRIS point cloud corner detection algorithm to extract the spatial corner data set in the corrugated plate weld point cloud data.Using the RANSCA polyline fitting algorithm to extract the weld trajectory and polyline angle size from the corner data set,and estimate the space trajectory normal vector to identify the weld position and posture,so as to predict the corrugated plate weld trajectory and the welding gun attitude recognition.3)Using swing arc to correct large deviation of corrugated plate welding seam.Since the corrugated sheet belongs to the thin plate category,thermal deformation is easy to occur during welding.Therefore,in order to improve the overall tracking accuracy of the weld,an arc sensor is used to correct large deviations in the welding process in real time.4)Verify the implementation scheme of this paper by building a corrugated plate welding seam tracking experimental platform.The designed corrugated plate welding seam tracking platform is used to perform trajectory recognition,welding torch posture adjustment and welding seam tracking experiment on the broken line welding seam,to verify the feasibility of the experimental method in this paper.