Research On Robot Deburring Position Compensation And Shape Measurement Technology

Casting products are widely used in industrial production and occupy an important position.However,due to mold manufacturing and casting processes,burrs are easily generated at the edges of the convex and concave parts of the cast parts and in the holes.The burr not only reduces the positioning accuracy of the workpiece to be processed,but also seriously affects the service life of the mechanical equipment.Domestic cleaning of burrs is mostly done by manual cleaning,with low processing efficiency and poor quality consistency.At present,the burr cleaning based on industrial robots has been put into use,and the related technology has made certain progress,but the absolute accuracy of the robot is not high,the position of the workpiece clamping point and the uncertainty of the contour shape,resulting in the integrity of the robot burr cleaning and The quality has not yet reached expectations.Aiming at the above problems,the research on the accurate measurement and error compensation technology of robot pose was carried out.At the same time,the threedimensional shape measurement of the workpiece and the correction method of the pose deviation were discussed.Combined with the relevant experiments,the above method was verified and analyzed.The main contents are as follows1)For the problem of low efficiency of robot pose measurement,the robot pose measurement method based on matrix singular value decomposition is studied: the coordinates of the target ball in the world coordinate system and the tool coordinate system are measured by using the laser tracker to solve the robot.The pose information,related experiments show that the measurement accuracy of the proposed method is close to the projection method,but the measurement efficiency is improved;2)Aiming at the problem that the absolute positioning accuracy of the robot is not high,a pose error compensation method based on the extreme learning machine is proposed: pose error and theoretical pose as the training data of the learning machine,and the regression model is constructed by the method of extreme learning machine.The pose error is predicted,and the pose error is compensated by modifying the pose pose.The experimental data comparison shows that the proposed method effectively improves the pose accuracy of the robot.3)The uncertainty of workpiece clamping point position and contour topography directly affects the deburring precision of the robot.The research on the 3D shape measurement and posture correction of the workpiece is carried out: the multi-angle of the workpiece is based on the Tianyuan 3D scanner.Scanning,point cloud stitching reconstructs the overall shape of the workpiece;the point cloud model registration method based on ICP algorithm is used to solve the workpiece pose deviation;the robot hand eye calibration realizes the transformation of the workpiece pose deviation to the robot world coordinate system,and optimizes the processing trajectory..