Research On Robot Control Method Based On Joint Posture

The traditional robot control method must obtain the target angle of each joint by inverse kinematics.The inverse kinematics solution can be divided into analytical method and numerical method.The analytical method requires the configuration of the robot to meet special criterion.The calculation amount of the numerical method increases in the form of geometric series as the degree of freedom of the robot increases.Aiming at the robot motion control method restricted by inverse kinematics,a robot control method based on joint posture is proposed in this thesis.The strategy is to continuously adjust the angle between the vector formed by the joint and the end-effector and the vector formed by the joint and the target position by getting the joint posture during the movement process.Because it is not necessary to solve the inverse kinematics,the matrix computation of the whole process is less.Compared with traditional method,this method has advantages in simplicity and universality.The main contents of the study are summarized as follows:(1)The principle of the robot control method based on joint posture on the rotational joint,prismatic joint and manipulator is studied.The random strategy is adopted to deal with the singular positions that may be encountered in the process of motion.Through the robot dynamics simulation platform Copplia Sim,the effectiveness of this method in planar linkage,manipulator and mobile manipulator is verified.Jaco arm is used to verify the proposed method.(2)The PID,computed torque and adaptive control law based on this method are designed.In the analysis of PID control,a simplified model analysis method is used.A robot adaptive control method with torque compensation is proposed to solve the problem of inaccurate robot dynamic model parameters.The adaptive control law is compared and analyzed with other control laws in the simulation environment.(3)To solve the problem of controller parameter optimization,a controller parameter optimization method based on FORL algorithm is proposed.In the simulation environment,the proposed optimization method is compared with method based on GA algorithm and PSO algorithm.The controller parameter optimization execution system is designed to improve the optimization efficiency.