Research On Robot Posture Control Method Based On Parallel CCD

When controlling a robotic arm,an inverse kinematic solution is usually required to find the angle of rotation of each joint of the arm to reach the target position.For multi-degree-of-freedom robotic arms,traditional inverse kinematics methods such as Jacobi matrix method and Newton’s iterative method have the drawbacks of large computational volume and poor real-time performance,while the analytical solution method has special requirements on the geometric configuration of the robotic arm and is not very versatile.In this paper,we propose a parallel CCD-based robotic arm posture control method,which firstly obtains the Cartesian coordinates of each joint of the robotic arm in real time by means of positive kinematics or position sensors,calculates the vector angle formed by the joint and end-effector as well as the joint and the target point,and then makes the joints of the robotic arm adjust the angle in real time during the motion.The angle is then adjusted in real time by the joints of the arm as it moves,so that the angle converges to zero and the end-effector reaches the target position with the specified attitude.Compared with the traditional inverse kinematic method,the proposed method in this paper has the advantages of low computational effort,high real-time performance and good generality.The main research of this paper is as follows:(1)Aiming at the problems of large computation and poor real time of robot inverse kinematics solution,a parallel CCD-based robotic arm posture control method is proposed on the basis of CCD algorithm.The principles of the proposed method on end-position control and end-posture control are introduced in detail,a random strategy is proposed to get rid of singular positions,joint motion constraints are added by setting upper and lower limits of joint angles and k-factor variables,and a Newton iteration method is combined to improve the convergence speed.(2)The design of a parallel CCD-based robotic arm posture control strategy is presented.Firstly,for the end position control problem,a single-task PID control strategy is proposed,and then a Lyapunov function is constructed to prove the stability of the control and derive the stability conditions of the control system.Then,a multi-objective particle swarm-based robotic arm posture control strategy is proposed for the end posture control problem,and the performance of the proposed control strategy is compared and analysed.(3)Simulation and physical experiments are designed to verify the advantages of the parallel CCD algorithm proposed in this paper.In the simulation environment,the proposed method is compared with the Jacobi matrix-based method to verify the superiority of the proposed method in terms of excute time,and the applicability of the proposed method to different robotic arms is verified by simulating a six-degree-of-freedom robotic arm,a ten-degree-of-freedom robotic arm and a mobile robotic arm.Experiments are carried out on the x Arm,Jaco and mobile robotic arm in a real environment to verify the advantages of the proposed method in terms of real-time,control accuracy applicability and robustness.