Manipulator Obstacle Avoidance Path Planning And Visual Servo Grasping

Random grasping is a classic problem in the operation of the manipulator.It is also a common and important task in modern industrial production.Recognizing and grasping disorderly-placed objects is a very challenging problem.The manipulator cannot rely on a set program to perform its work,but needs to perceive and analyze the environment,and then make judgments to plan and control.To control the manipulator to complete random-grasping tasks,we must first plan a safe and short path,then control the manipulator to arrive an accurate pre-grasping posture.If there are some obstacles in the workspace,the path of the manipulator needs to avoid these obstacles.Therefore,aiming at the manipulator grasping task,this paper solves the problem of controlling the manipulator to grasp the target workpiece in the environment with obstacles,mainly including obstacle-avoidance path planning and visual servo control.The main research work of this paper is as follows:(1)Plan the obstacle-avoidance path for the manipulator.Improve the traditional RRT*path planning algorithm in consideration of its shortcomings,and extend the algorithm to the joint space of the manipulator:·In consideration of the long time of traditional RRT*planning,the method of using Bezier curve to optimize the path is proposed.This algorithm can get a shorter collision-free path in a less period of time·To combine the obstacle-avoidance algorithm with the manipulator model better,randomly sample in the joint space of the manipulator and then plan the path,so that it is not necessary to solve much inverse kinematics during the planning process,which will reduce the amount of calculation.·The simulation experiment with the 6-DOF manipulator proves the safety and feasibility of the path planning algorithm proposed in this paper.(2)Visual servo control of the manipulator.According to the shortcomings of the traditional IB VS algorithm,an improved IB VS algorithm is proposed and achieves good controlling effects:·In traditional IBVS algorithm,when the initial feature error is large,the linear velocity along the z-axis of the camera coordinate and the angular velocity around the z-axis are close to zero.Therefore,this paper proposes to add two control components·The comparative simulation experiment with a 6-DOF manipulator proves the effectiveness and feasibility of the improved visual servo algorithm(3)Carry out the experiment of manipulator grasping.Combine the path planning algorithm with the visual servo algorithm proposed in this paper,and control the manipulator to complete the task of grasping the target workpiece in an obstacle environment:·Use the obstacle-avoidance path planning algorithm to control the manipulator to reach the target workpiece,then use the visual servo algorithm to control the manipulator to reach the accurate grasping position and posture.Grasp the workpiece and place it at a set position.The experiment result shows the algorithm proposed is practical in real manipulator working scene.