Research On Improving The Grasping Accuracy Of 6-DOF Series Manipulator

Since the 20 th century,robot technology has been developed rapidly,especially in recent years,with the development of computer and related technology,intelligent robot has become a hot topic of human research.As a very important part of robot,manipulator has been applied in many occasions,such as factory,hospital,school,etc.With the increasing application of the manipulator,how to improve the grasping accuracy of the manipulator has become a key technical problem to be solved.At present,the 6-DOF series manipulator with visual servo is the most widely used grabbing system,which can improve the grabbing accuracy of the system mainly from the control accuracy of the manipulator and the accuracy of visual information processing.In order to eliminate the steady-state error caused by gravity,the realtime gravity compensation and fixed gravity compensation are carried out for the system without load and with load.In order to realize the real-time gravity compensation of the system,a parameter fitting method based on neural network is proposed.The data of the influence of the gravity of the manipulator itself and the load on each joint of the manipulator are fitted with the neural network,and the selftrained network model is used in the real-time compensation of gravity in the system.The simulation results show that the proposed method can effectively compensate the gravity of the system and improve the control accuracy.In view of the influence of the system inertia on the control accuracy of the manipulator without load and with load,a control strategy based on real-time compensation system inertia parameters is proposed to improve the control accuracy of the system.The neural network is used to fit the real-time inertia parameters of the system,and the trained network is used to compensate the real-time inertia parameters of the system.Through the simulation experiment,it is verified that the proposed method can improve the control accuracy of the system.At last,aiming at the part of vision recognition detection,this paper presents the detection and matching of feature points based on SURF algorithm,and improves the matching results by combining GMS algorithm and RANSAC algorithm to improve the accuracy of image matching,and the method based on contour extraction is used to estimate the pose of the target object,so as to improve the recognition accuracy of the object to be grasped and get its posture.The validity of the proposed method is proved by comparative experiments.