Trajectory Planning and Minimum-Norm Invers e Kinematics Solution for a Humanoid Redundant Arm

In this paper, trajectory generation in task space and minimum-norm inverse kinematics solution for a humanoid redundant arm is discussed. To this end, in order to state desired task for the end point of the arm, cubic polynomials of time for the trajectories in Cartesian s pace are generated. Due to redundancy of the arm, various solutions for the invers e kinematics based on available constraints and task priorities may be assumed. In order to exploit the redundancy of the arm to generate motion with minimum norm of control effort, Moore - Penrose Invers e (Pseudo invers e) is adopted. Using this method, the desired velocities of the joint s pace variables are encapsulated. These values are employed as desired values in the velocity control procedure to generate motion that is locally minimized. Effectiveness s of the proposed method is shown in simulation. Through implementation on an experimental test bed that is developed at the Center of Advanced Systems and Technologies (CAST), results are discussed and verified.