Research On Vision-guided Dual Arm Collaboration In Module Intelligent Assembly System

With the development of science and technology,the structure of industrial products is becoming more and more diversified.The rich and varied processing and manufacturing methods have also put more complex functional requirements on industrial robots.Machine vision technology has the advantages of real-time,non-contact,high accuracy,low cost,and high digitization.It can be combined with industrial robots,which is particularly good at flexible production and assembly processes of parts with complex structures and variable shape or position.This production method is rapidly spreading in industrial area.Based on the actual needs of enterprises,this paper takes the street light module as the application object,researches the vision-guided dual-arm coordinated automatic assembly technology,and completes the development of the vision-guided dual-arm coordinated module intelligent assembly system.The main contents include:(1)The research status of vision-guided industrial robot assembly technology is analyzed,then the functional requirements of the module intelligent assembly system and the system design plan are given,and the system hardware selection and the design of the end effector are completed.(2)According to the task division of the dual-arm system,this paper developed cooperative motion strategy for two robot arms.And according to the requirements of the strategy,the calibration of the two bases coordinate was completed to realize the transformation of the space pose between the two coordinate systems.Aiming at the problem that the arms are prone to collision in a small space during assembly,a pose constraint method based on the virtual coordinate system is proposed,which can establish the position constraint relationship between the two-arm tools.This method effectively implements the follow-up control of the arms to avoid collisions.Then,in this paper,the forward and inverse kinematics of the two arms were derived using the D-H method,and the dual kinematics were simulated using MATLAB to verify the effectiveness of the pose constraint method.(3)In the visual guidance module,the auxiliary coordinate system calibration method was used to obtain the position conversion matrix from the image coordinate system to the robot base coordinate system.Aiming at the problem of inconsistent gray values of the image that caused by uncertain changes in ambient light,this paper proposed an adaptive threshold segmentation method based on image features that binarized the image effectively;And combined with image filtering,Huff transform and other methods,a complete and robust image processing algorithm flow is completed,which can accurate extraction of goals.(4)In this paper,a vision-guided dual-arm collaborative module intelligent assembly prototype system was developed for testing and analysis of system errors;system control software was developed using Qt.Then,the entire system was tested for operation,which verified the effectiveness and stability of the system.