Font Size: a A A

Design And Implementation Of Coordinated Control Algorithm For Dual Amnipulator Based On Binocular Vision

Posted on:2020-12-26Degree:MasterType:Thesis
Country:ChinaCandidate:Q TangFull Text:PDF
GTID:2428330596992797Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
This paper studies the camera calibration,digital image processing,positioning and grabbing of work pieces,digital signal processing,and closed-loop control strategy based on force sensor in these hardware of dual-arm robot,binocular camera and six-dimensional force sensor.The two-arm coordinated control algorithm under the guidance of visual positioning and real-time feedback of force signals is designed to control the two-arm robot to coordinate the screw assembly task.Detecting the pose of the workpiece based on monocular vision.The Canny operator is used to extract the contour of the target workpiece in the image.The position of the workpiece is determined by determining the centroid.The top and bottom circles of the workpiece are fitted by the Imfindcircles function method and the Hough transform method respectively.Two centers are obtained by fitting the result.combined with the workpiece to obtain the attitude of the workpiece.The position of the workpiece is detected based on binocular vision.On the basis of the first stage,combined with morphological processing and image matrix operation,and based on the characteristics of the workpiece,a feature point extraction algorithm based on the feature line in the target geometry is designed.The feature points are obtained by matching the feature points in the left and right images.The coordinates in the three-dimensional space,and then the pose of the workpiece.Design a single robotic arm assembly screw algorithm.Use binocular vision to position the nut on any position on the desktop,control the arm grabbing,position the screw placed horizontally on the table,control the movement of the arm to make the nut touch the screw;study the closed-loop control algorithm based on the real-time feedback data of the force sensor,algorithm It is applied to the following steps: workpiece contact judgment,screw fastening,screwing in and verification task is completed.Design the double robotic arm assembly screw algorithm.Based on the algorithm and process design involved in the single robotic arm,the basic idea of the algorithm is unchanged,and the coordinate system needs to be transplanted into the dual-arm space.The complete steps include: visual positioning of the nut and grasping,arms movement to the pre-contact position,main arm Adjust the posture from the arm movement until the contact,the screw buckle,and the screw into the screw to verify whether the task is completed.Based on the existing control platform,the code is written,improved and added to realize the control strategy designed in this paper.The dual-arm robot experimental platform is used to verify the two-arm coordinated control algorithm,which proves the feasibility of the algorithm and control system.
Keywords/Search Tags:binocular vision, six-dimensional force sensor, image processing, coordination control algorithm
PDF Full Text Request
Related items