Analysis and control of power grasping

The problem of grasping objects with a robotic hand is considered. The possibility that some or all of the fingers are not able to arbitrarily control interactions with the grasped objects is taken into account. Such defective manipulation systems can still be cooperatively coordinated so as to perform usefully. In fact, having defective arms is the norm rather than an exception in many manipulation operations, such as power grasps with a hand, or whole arm manipulation of large objects. The paper attempts to solve the problem of optimizing contact forces in the power grasp of an object. To do so, a basis of the subspace of grasp forces that are available for grasp optimization is firstly established. This analysis is instrumental for subsequent optimization strategies, incorporating the quest for an extremum of some quality criterion. A control algorithm is presented that guarantees the asymptotic convergence to the grasp force configuration that minimizes the risk of slippage while maintaining bounded contact forces between the fingers and the object.<<ETX>>