Experimental Research On Industrial Robot Processing System Based On Machine Vision

With the continuous development of modern science and technology,it is inevitable for industrial robots to replace human operations in factories.The efficiency of industrial robot is higher than that of human,but it can only work according to the pre-set program flow,If the position of the object to be processed changes,or the processing environment changes,the industrial robot cannot continue to work according to the original program,so it needs to manually rewrite the control program.Therefore,it is necessary to install "eyes" for industrial robots.It is necessary to combine machine vision technology with robots,guide robots to complete the work of unknown target by machine vision,collect the image of the processed workpiece by the industrial camera of the processing system,and further process the information data of the image to obtain the position of the preset feature points in the image The image model of industrial camera is used to convert the pixel coordinates into the coordinate data under the world coordinate system.The coordinate information is transmitted through Ethernet.The robot controller analyzes the data.The inverse kinematics solution of industrial robot is used to solve the angle change of each joint of industrial robot and generate the processing path.The industrial robot drives the pneumatic cutter to carry out the target workpiece according to the instructions processing.The main hardware structure of the system can be divided into industrial robot control system and image acquisition system.This paper analyzes the different categories of machine vision and control industrial robots,and selects the Eye-to-Hand end open-loop scheme of position control based on fixed monocular vision.The hardware and equipment involved are analyzed and selected,the realization process of software flow is analyzed,and the communication of the whole system is established.The model of machine vision guided industrial robot machining is established,and the control strategy of machine vision guided industrial robot machining is set up.The pose and coordinate system of industrial robot are described.The kinematics model of MISUBISHI six axis industrial robot is established by D-H modeling method,and the inverse kinematics of the robot is calculated to ensure the pose relationship between the end effector and the workbench of the industrial r obot,To ensure that the industrial robot will not collide with other parts on the workbench during the operation process,the working path of the platform is simulated by MELFA-Works control to ensure that there is no singular point in the operation of the industrial robot.Using Vision Master algorithm platform to filter and binarize the acquired image,the important information in the image is 20enhanced.In this paper,local feature point template matching is used to identify and locate,local feature template is set,fast feature extraction is carried out,pixel information of feature points is obtained,industrial camera is calibrated,image generation principle of industrial array camera in the system is analyzed,and mathematical model of image generation of industrial array camera in the system is obtained,pixel coordinates and Based on the calculation formula of world coordinate.An experimental platform was built to determine the tool coordinate system of the end effector of the industrial robot,and the processing parameters were measured at the same time.Through the camera system and industrial robot to complete the system calibration,the pixel coordinates are transformed into the world coordinates,and the processing program is compiled with MELFA-BASIC V language.The information sent by the industrial camera is analyzed,and then the processing experiment of the work piece is carried out,and the error is analyzed.The experimental results show that the error between the calculated coordinate and the preset coordinate is within 0.3mm,and the qualified rate is 97.18%.The system can process the target workpiece accurately and realize the automatic processing of industrial robot guided by machine vision.