Dynamic Parameter Identification And Trajectory Planning Of A 6-DOG Humanoid Manipulator

The application of robotic arms in the industrial field has become increasingly widespread,and the identification of dynamic parameters and optimal trajectory planning for robotic arms remains a hot research topic.This paper focuses on the six-degree-of-freedom humanoid robotic arm built in the laboratory,and conducts research on its kinematics,dynamics,identification of dynamic parameters,and optimal trajectory planning.Firstly,the robotic arm platform is built and its simulation model is established.The D-H method is used to analyze its kinematics,establish corresponding coordinate systems,and calculate the transformation matrix between each joint to obtain the forward kinematic model.Inverse kinematics is studied through numerical calculations to accurately calculate the angle of each joint.Experimental simulations in MATLAB are carried out to verify its correctness.The dynamic model is solved using the Newton-Euler iteration method,and is transformed into a linear form for subsequent parameter identification work.Then,the corresponding observation matrix and minimum inertial parameter set are obtained.Secondly,an offline identification experiment of dynamic parameters using the global identification method is designed.A trajectory is designed based on the physical characteristics of the robotic arm,such as velocity and acceleration,using the Fourier series as an excitation trajectory.The trajectory can effectively excite the characteristics of various dynamic parameters of the robotic arm.Then,the robotic arm runs along the trajectory,and the position,velocity,and torque of each joint are collected.Based on the filtered data using the Kalman filter,an improved iterative reweighted least squares method is used for parameter identification,and the accuracy of the parameters is experimentally verified.Finally,based on the identified dynamic model,time-optimal trajectory planning with dynamic constraints is studied.In order to solve the local and global optimal problems of the general time-optimal problem under dynamic constraints,it is transformed into a convex optimization control model,and the physical and dynamic constraints of this robotic arm are added.Under the time-optimal objective,the factors influencing the torque change rate are considered,so that the robotic arm can be more easily executed.The feasibility of the algorithm is verified through experiments.