Researching On Recognition And Trajectory Planning On Space Curve Weld-Seam By Robot

Automation and intelligent welding process is the future trend of welding technology.The usage of welding robot id key method to make that come true.However,mostof the using welding robots follow teaching-playback style,which has low intelligence and adaptive ability.Seam tracking technology is an important part of intelligent welding,which makes robots adapt to unknown and variable environments by using the multiple sensors.In this paper,we focus on information acquisition and trajectory planning of 3-D seams which has additional height differences.The passive vision is used to recognize the initial point of the seam and acquire its 3-D information while the active vision is used to recognize the target point and its 3-D information during seam tracking.First of all,the hardware system of the passive and active vision eye-in-hand robot is built.The system includes the robot model and control unit,PC camera and linear laser device used in active vision.The camera is calibrated by the Matlab Calibtoolbox and the checkerboard to obtain its intrinsic parameters,distortion parameters and the extrinsic parameters.In passive vision we use motor algebra to calibrate the eye-in-hand system,the algorithm is more accurate and faster than the traditional algorithm.The morphologymethod is used to detect the edges of the image,the adaptive ability of the method is better than other traditional operators;than Hough transform is used to detect lines in the image to figure out the initial point.When recognizing the target point in tracking process,the welding seams and the laser bars are separated by threshold which is obtained according to the features of the images;then a skeletonization method based on maximum disc is used to thinning the laser and the seam streaks;at last the Hough transform is used to figure out the target point.An ‘one eye with two positions' method,which achieved by moving the robot,is employed to obtain the 3-D information of the initial point.In tracking process,we use the linear structure light 3-D measurement model to calculate 3-D information of target point.Finally,a trajectory planningand posture transformation method based on short lines is proposed and a servo control system is eventually designed.The experiments show that the proposed system satisfies the real application with high precision and good tracking result.