Research And Development Of Robot Welding Seam Tracking Control Technology Based On Visual Guide

As the rapid development of the computer technology and artificial intelligence andso,the demands for industrial automation and intelligent level of welding technology continueto increase.In order to improve the factory efficiency and the quality of welding operation,structure light vision is used to solve the problem of welding robot during weld automatictracking.Seam tracking control of structure light vision guided robot,is a process to completethe welding operation based on structure light vision perception of weld to forecast planningout the precise trajectory of welding torch under the servo control.Aiming at the problem of existence of sliding mode control of the robot by larger initialposition deviation caused by big torque and the speed jump, double fuzzy adaptive slidingmode control algorithm is proposed.Using an adaptive fuzzy controller, according to theconditions on the sliding mode reaching slide switch gain estimate, eliminating thephenomenon of output torque chattering in the sliding mode control,so enhancing its ability toadapt to the uncertainty factors.Also,using another adaptive fuzzy controller for correction thecoefficient of exponential reaching law.Experimental results show that this method improvedthe large torque and the speed jump problems caused the deviation due to larger initialposition.As for problems of failure access to weld the initial points of3D information of thestructure light vision through image processing, using a laser scanning method of weldpositioning initial point.Making the straight line direction perpendicular to the laser line asrobot translation direction,while improve the scanning speed also increasing initial pointpositioning accuracy.In the process of translation control shaft rotation by the lower end of thestructure of fuzzy controller, guaranteeing the weld point image coordinates in the expectedimage coordinates of the acceptable range.The experimental results show that the method canquickly and accurately position the initial weld point, and to ensure that the continuous weldimage.In view of the incongruity between the visual tracking and the welding robot movement,using strategy of the welding robot movement decomposition into rotating part and translationpart.Translational motion guarantees the precision of welding torch,while rotating guaranteesgood welding robot posture.So using visual servo control feedback the visual images, andwelding torch coordinate system modeling method, robot real-time adjustment.Experimentsshow that the strategy under the condition of the weld tracking accuracy is not affected, whileincreasing the degree of seam tracking position complaisant and continuity of tracking.Finally, using a delayed tracking strategy to track the seam for a given weldingseam.Control program is written in VS2010platform, the experiments proved that weldingtorch can be automatic positioned to weld the initial points, and tracking the weldautomaticly.Ability to obtain continuous images in the process of tracking for real-timeadjustment of posture, to ensure that the tracking is completed and the seam is trackedaccurately by the torch.