Hand-Eye Coordination Research For Robots For Industrial Assembly Applications

With the coming of the age of intelligence,intelligent industrial robots integrated with vision system have penetrated into many fields.In particular,robots with hand-eye functions play an important role in the fields of automobile manufacturing,logistics,and aviation exploration.In order to meet the needs of flexible production line to quickly locate the target and achieve the specific tasks,this paper studied the problem of automatic assembly in industrial production line based on the hand-eye coordination method of industrial robot.The main contents of this project are as follows:Firstly,in order to meet the needs of flexible assembly on industrial production lines,system design and component selection were completed according to product specifications.The system was divided into a series of static visual feedback control scenes of image processing and robot control,and a dynamic visual feedback scene of image processing and robot control parallel,it also introduced the working principle of the two scenarios.Secondly,the hand-eye system scheme of the monocular Eye-in-Hand was determined,and the hand-eye coordination based on the uncalibrated technique was used for the static visual feedback control scene to complete the task of automated assembly;Hand-eye coordination based on calibration technology was used for dynamic visual feedback control scene to complete the task of automatic assembly.The positioning accuracy of the system in two scenarios was analyzed by experimental data.Thirdly,the background difference method,Otsu thresholding and other morphological analysis methods were used to obtain the target contour.A rectangular fitting algorithm was introduced to calculate the centroid information of the component image to achieve positioning,and a frame difference method was introduced to calculate the motion speed of the product.Fourthly,in the static and dynamic visual feedback scenarios,the grab assembly experiments were carried out,the corresponding experimental data were recorded,the influence of illumination intensity and speed on the dynamic scene assembly results was analyzed,and the system uncertainty was analyzed.Experiments show that the target assembly success rate of the two visual feedback scenarios is 100%,which can meet the requirements of industrial production lines.