Research And Development On Cement Bag Loading Robot

In most medium and small sized cement factory,cement bags are basically relied on manual operations currently,and many solid dusts are inhaled by workers during manual operations,which increases the incidence of pneumoconiosis and seriously threatens health of people.In response to these problems,this project has designed and developed an industrial robot that automates the loading of cement bags under the guidance of machine vision.The main contents of the research include the mechanical system design of cement bag loading robot,the design of motion control systems,and the transporter carriage recognition system.Firstly,on the basis of investigation and comparison of structures of common palletizing robots,According to the demand of cement bag loading,a mechanical structure of cartesian coordinate loading robot is put forward.According to the working parameters and the dimensions of transporter carriage,the mechanical structure of loading robot was designed with four degrees of freedom.The servo motor was used to drive each motion axis and the cylinder was used to drive the cement bag gripper.By the dynamic simulation analysis of motion mechanism,the required power of each moving axis is obtained,which validates reasonableness of mechanism motion.By the finite element analysis of the key mechanism of the loading robot,it is proved that designed mechanical structure meets the requirements of use.According to the characteristics of four motion axes of the loading robot,this article designs motion control network based on the Ethernet bus and CANopen bus.The core component of hardware system adopts the industrial-grade and multi-function motion controller DVP10MC11 T,which constitutes the CANopen network with servo drivers.By using the method of electronic gears to achieve the synchronous movement of two synchronous servo motors in X direction,in order to ensure the safety of loading robot,The close switches are adopted as limit switches of the linear axes.The software system mainly includes the motion control program design and the human-machine interface design.The motion control program completes the control of multiple motion axes in the CANopen network by the DS301 protocol.According to the imaging target and the characteristics of working area,the imaging system is designed.The model of industrial camera,industrial lens and lighting source is selected,vehicle recognition based on shape detection is realized.The calibration of the imaging system is carried out by point distance calibration,perspective calibration and distortion model calibration.The camera coordinate system and the robot coordinate system are calibrated by adjusting the camera position and the transformation of the pixel coordinates with the robot coordinates.Motion control and vehicle recognition experiment were carried out on the Longmen test platform,which verified the rationality of the motion control system and vehicle recognition system design.