Research On Vision Calibration And Motion Control In Automatic Container Unit For Mold Processing Mold Automation Scheme

With the development of China's manufacturing industry and the decline of the demographic dividend,coupled with the government's strong support in recent years,the development of the manufacturing industry towards automation and intelligence is the general trend.This article focuses on some precision and efficiency problems we encountered in the implementation of mold automation solutions and do some research on them.First,we introduce the working principles of the two research objects,i.e.,the automatic container unit and the SCARA robot.It mainly introduces the working principle of the automatic container unit,the structural design and layout of the silo holes,the structure of the robot and the electrical selection.Secondly,in order to solve the problem that the absolute positioning accuracy of a SCARA robot is not good enough to meet the needs of the project,a scheme for calibrating the key kinematic parameters of a SCARA robot using a single camera is proposed,and the rotation of the dot calibration plate is used to fit the flange center of the robot.,Using SVD to solve the system of statically indefinite homogeneous equations,fitting three key kinematic parameters.The maximum deviation of the actual trajectory of the calibrated robot when performing 70 mm linear interpolation is less than 0.20 mm.Thirdly,in view of that there are endlessly repetitive calibrations jobs in the implementation of the mold automation solution,a binocular vision-based bin hole calibration method was proposed and applied to automatic container unit products.Firstly,a binocular vision-assisted calibration tool was designed according to requirements.In the case that the feature points cannot be found through general image processing algorithms,the local template is used to estimate the affine transformation of the overall template,and the prior knowledge is used to predict the key feature points.Finally,the operation process and experimental results of calibrating the hole position of the bin using this tool are introduced.The final conclusion is that the binocular visionassisted calibration tool can achieve an accuracy of 0.20 mm in the XY direction and 0.30 mm in the Z direction,and the recognition rate of the silo hole position can reach 93.33%,which can shorten the time required for calibration job from a day shift to 70 minutes.Finally,this article introduces the design and implementation process of the motion control library of the multi-joint robot control software The motion control requirements of different robots or motion mechanisms are summarized into a set of basic APIs,and the basic instructions of different motion controllers are used to implement these interfaces.The development of the motion control library aims to improve the efficiency of development and maintenance of the automation equipment software.In the end,the motion control library was applied to the SCARA robot and the automatic container unit.