Robot motion generation considering external and internal impulses

摘要

This paper deals with motion generation algorithm considering the external and internal impulses. Initially, we analyze the trend of the impulses with a simple 3-DOF planar robot arm in various contact positions and present a problem that is in discord with the general trend. To cope with this problem, we propose a new motion generation algorithm considering both the external and internal impulse. The gradient projection method is employed to exploit the kinematic redundancy of robot systems. Initially, the effectiveness of the proposed algorithms is verified through simulation of a 3-DOF planar robot arm model, and two application examples considering both the external and internal impulses are investigated. Through simulations for the landing of a 4-DOF planar human-body model and the sawing motion by a dual-arm model, it is shown that the posture is stable and the trend of the impulses calculated by the analytical models is coincident to the human experiences.