Visual Inspection And Sorting Of Fasteners Based On AGV

In recent years,with the in-depth promotion of "Industry 4.0",which focuses on intelligent production and intelligent logistics,intelligent sorting robots are playing an increasingly important role in the fields of logistics,workpiece sorting,and so on.However,with the rapid expansion of manufacturing industries such as machinery and automobiles,large quantities of mechanical parts have been mixed together.How to intelligently sort and flexibly transport parts has become an urgent issue for major manufacturers.In response to the above situation,this article has independently designed a sorting type AGV,which is based on the traditional AGV and carries a manipulator with existing sorting functions.It mainly includes autonomous navigation,target recognition and positioning,intelligent sorting,and other functions.The main research content of this article includes the following aspects:(1)Firstly,the mechanical design of the AGV and robotic arm was designed and completed.The AGV adopts a six wheel differential steering chassis structure,and two drive wheels and four universal wheels are installed on the chassis.At the same time,multiple sensors are fused to achieve AGV obstacle avoidance,docking,and other functions;Sorting manipulator selects three-axis manipulator,and forms sorting AGV together with AGV.Secondly,the AGV motion model is analyzed.Finally,the forward and inverse kinematics of the manipulator is analyzed by D-H modeling method.(2)To solve the problem of autonomous navigation of AGV,visual tracing is selected as its guidance mode.According to the movement characteristics of AGVs on different sections,threshold control and fuzzy control are combined to meet the requirements of trajectory following and flexible transportation.(3)The research on sorting and grasping objects based on binocular vision manipulator is carried out,and the methods of detecting and positioning objects are introduced.Firstly,the visual model related to robot grasping objects is introduced,and then the YOLOv5 s target detection algorithm is selected to identify objects.Finally,in terms of binocular vision positioning,the advantages and disadvantages of several common stereo matching algorithms are analyzed,and the BM stereo matching algorithm is selected to meet the real-time sorting and positioning accuracy requirements of the robot in sorting objects.Based on the above proposed functions,on-site experiments and tests were conducted on the independently designed sorting type AGV.The experimental results show that the AGV has good track following ability in different road sections,and the average recognition rate and sorting success rate can well meet actual requirements in sorting fasteners.