Research On Visual Positioning And Control Technology Of Industrial Robots For Circular Hole Measurement

In the nuclear industry,the measurement of the geometric parameters of the deep bore of the heater tube sheet is a complex and arduous task,which is currently measured manually using contact gauge tools.Loading the contact force sensor at the end of the robot instead of manually measuring can greatly improve the measurement efficiency,but in the application,the posture of the workpiece needs to be obtained through teaching.When there is a large manufacturing error in the workpiece or a deviation occurs in the workpiece clamping position,the position of the circular hole of the workpiece is different from the teaching position,which easily causes the robot to collide with the workpiece and damage the sensor.To solve this problem,this article uses a vision-guided robot to automatically approach the workpiece for the measurement tasks.Firstly,the internal parameters of the camera and the relative positional relationship between the camera and the robot are obtained through visual calibration.The image processing algorithm is then used to extract features of the workpiece image,and visual positioning is implemented according to the image features.Finally,the workpiece position information obtained by visual positioning or directly controls the movement of the robot to the workpiece according to the image information.The main work of the thesis is as follows: 1.An improved Tsai-Lenz hand-eye calibration method based on least squares method and data filtration is proposed to complete the hand-eye calibration of the robot vision system,and the accuracy of the hand-eye calibration is tested by using the robot end tool to contact the checkerboard corner point measurement method.Analyzed the circular hole workpiece,and used the maximum inter-class segmentation method to obtain the threshold value.The workpiece image is extracted from the image by threshold segmentation,and then the extracted image edge pixel points are fitted into an ellipse using a least square method.At the same time,the four vertices of the ellipse are used as feature points,and the positioning of the target workpiece is realized according to the imaging principle of the camera.Then through the kinematics and space coordinate transformation of the robot,the pose of the workpiece under the camera coordinates is mapped into the robot base coordinate system,the simulation of visual control robot to workpiece motion is realized.2.The process of moving the end of the robot is simplified into two-plane rotation transformation and plane translation transformation.According to the axis-angle relationship when the plane rotates,calculate the position and orientation of the robot when it reaches the measurement position.And construct a position and orientation presented robot vision control system,and then test the vision-guided robot approaching the circular hole workpiece,and verified the feasibility of the method from the perspective of the positioning accuracy of the robot end probe.3.Selected the moment of the image is as the image feature,and the PID controller is constructed to realize the image-based robot visual servo control system to realize the simulation analysis of the system.And built a robot vision guidance system,and a visionguided robot system software approaches a circular hole in the workpiece.