Computing of the sub-optimal grasping forces for manipulation of a rough object by multifingered robot hand

Abstract

This paper proposes a new method for computing the grasping forces of object grasped by multiple cooperating manipulators or a multifingered robot hand. The intention is to find grasping forces that ensure safe grasp on the object during arbitrary manipulation. The grasp should be with minimal internal loading of the object. The finger/object contacts are treated as the surface contacts with Coulomb friction. Consequently, the contact friction moments are not being ignored. This method predicts the magnitude of the inwardly directed normal components of the grasping forces. These components and the tangential (friction) components of the grasping forces and grasping friction moments that act normally to the object at each finger/object interface hold the object. To solve the problem the limitations of the grasping force and moment components were defined. Force components were divided into active and passive components. The influences of the friction force components on the force and on the moment equilibrium were considered separately. The iterative algorithm for predicting grasping force and moment components, satisfying their limitations and condition of the minimal internal loading, is proposed.