Research On Design And Application Of Manipulator Servo Control System Based On Visual Feedback

In recent years,robot vision servo control technology has been widely used in various fields,which has become a hot spot in the research of robot based on visual closed-loop feedback control.In this paper,a monocular eye-in-hand robot is taken as the research object,combined with the visual feature tracking algorithm,an image-based and position-based closed-loop visual servo robot grasping system is designed.On this basis,in order to further solve the optimal problem of the robot’s motion path,a convex optimization solution,a visual servo control system based on second-order cone planning,is proposed.The main research contents of this paper are as follows.Firstly,according to the characteristics of the manipulator,the communication relationship between the upper computer and the manipulator is established.The relevant mathematical model of the visual servo system is constructed by the D-H parameters of the manipulator and the camera imaging theory.It includes robot linkage coordinate system,forward kinematics model,camera aperture imaging model and PNP pose estimation model.In the visual servo control system,in order to map the conversion relationship between manipulator joints and image information,the Jacobian matrix model in the system is given,which lays a theoretical foundation for the subsequent uncalibrated closed-loop visual servo system.Secondly,for the capture of different objects in multiple scenes,the corresponding detection and tracking algorithm needs to be selected according to the characteristics of the captured target.A stable tracking algorithm is used to obtain the characteristic information of the target object in real time.In the traditional tracking algorithm,image moment,mean shift algorithm and pyramid optical flow tracking algorithm are used to track different target objects;In the deep learning algorithm,combined with the Yolo framework,the training model is designed to realize the motion tracking of the robot grasping target.Further,in order to apply the above-mentioned image tracking algorithm to the robotic arm visual servo grasping system,an image-based visual servo control system(IBVS)and a positionbased visual servo control system(PBVS)are established,and the stability of the system is effectively analyzed.Through different experiments,the effectiveness of the two servo systems is proved,and it is proved that the image-based visual servo control system can effectively control the motion trajectory of image features in the two-dimensional image plane;The position based visual servo control system can effectively control the motion trajectory of camera pose in threedimensional spaceFinally,in order to comprehensively solve the optimal problem of two-dimensional plane image feature trajectory and three-dimensional space robot motion trajectory in the robot servo process,a closed-loop visual servo control system based on second-order cone programming is designed.the control system can realize the image at the same time effective control in space and the position space,and constraints on the image visual field,to ensure the servo process,The target object is always in a valid camera field of view,avoiding servo failure due to image loss.At the same time,the joint velocity and joint Angle of the robot are constrained to ensure that the robot keeps a reasonable speed in the servo process,and avoid the servo interruption caused by the loss of image features caused by speed mutation.Through experiments,it is proved that this scheme can effectively overcome the shortcomings of IBVS and PBVS,and effectively improve the robustness of servo control system.