Research On Assembly Technology Of Industrial Robot Based On Vision

Assembly is widely used in manufacturing industry.There are some problems in manual assembly,such as low assembly efficiency,low precision and high labor intensity.In recent years,the technology of industrial robot and machine vision has developed rapidly.Combining the two can greatly improve the intelligent level of industrial robot.In order to get rid of the boring assembly work and improve the precision and efficiency of assembly,this paper designs a set of vision based assembly system of industrial robot,which realizes the precise assembly.The main contents of this paper are as follows:Firstly,according to the properties of assembly work pieces and the requirements of this project,the overall scheme design of the assembly system of industrial robot based on vision is developed,which is divided into software system design and hardware system design.The hardware system design introduces the selection of Daming TM5 industrial robot,industrial camera,industrial lens and light source,and establishes the hardware platform of the assembly system.The software system design part completes the design and development of camera module,hand eye calibration module,template module,communication module and so on.Secondly,the industrial camera is calibrated and the hand eye calibration is carried out between the industrial robot and the industrial camera.In order to improve the assembly accuracy and assembly efficiency,this paper uses the combination of eye in hand and eye to hand to locate the target workpiece.The internal parameters and distortion coefficients of the industrial camera are calculated by using MATLAB software.The eye calibration of eye system is carried out by nine point calibration method.The eye calibration program of industrial robot is used to calibrate eye of eye system of eye in hand system.The conversion matrix between the coordinate system and pixel coordinate system of industrial robot is calculated.Then,the positioning and angle correction algorithm of the target workpiece are studied.This paper applies different preprocessing methods to the original image acquisition,and selects the suitable pretreatment method according to the processing results,which is convenient for the later image processing.This paper uses different positioning algorithms to locate the target workpiece,and according to the results of positioning,the positioning algorithm suitable for this topic is selected.Because the industrial robot will rotate the target workpiece in the process of grasping the target workpiece,this paper corrects the angle of the target workpiece after the grab.In this paper,the sub-pixel edge detection is used to find the rotation angle of the target workpiece.Finally,the assembly experiment of the designed assembly system of industrial robot based on vision is carried out,and the experimental results are analyzed.The assembly system of the industrial robot based on vision has good assembly effect by combining eye in hand and eye to hand and angle correction system based on sub-pixel to assembly the shaft hole.The positioning accuracy and assembly success rate meet the needs of this project.