Time Optimal Motion Planning and Admittance Control for Cooperative Grasping

Abstract

Cooperative grasping refers to the situation when an object is manipulated by multiple robots and the grasp is achieved by the unilateral contact between the robots and the object. This is different from the cooperation of multiple robots where each robot rigidly grasps the object. Motion planning of cooperative grasping tasks involves active force control of the interaction wrench in order to ensure stable grasp. This becomes particularly challenging when aiming at time optimal motions. It is crucial that the trajectories are continuous up to third-order, in order to satisfy velocity, acceleration, and jerk as well as torque limits of the robots. A solution approach is presented for the time optimal path following of two robots performing cooperative grasping tasks. The time optimal path is determined with a dynamic programming method. An admittance control scheme in task space is proposed and used to generate the contact wrench. The method is applicable to