Humanoid Robot Arm Kinematics Modeling and Motion Planning

Abstract

This paper establishes the kinematics model of humanoid robot arm, the arm forward kinematics equations were built and solved, based on the advantages of CCD (Cyclic Coordinate Descent) and BFS (Broyden-Fletcher-Shanno) algorithm to solve the inverse kinematics of humanoid robot arm. In the joint space, using cubic polynomial and quintic polynomial interpolation method respectively for each joint angle interpolation. Cubic polynomial trajectory planning can meet the point-to-point movement in general, but can not guarantee the continuity of acceleration of each point; Quintic polynomial trajectory planning can ensure that each point is continuous of the joint angle, angular velocity, and angular acceleration, so this polynomial method can meet the movement of the humanoid robot arm.