| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 411708 | Robotics and Autonomous Systems | 2009 | 9 Pages |
The capability to equilibrate external wrenches is crucial in optimal grasp planning. This paper presents a new method for evaluating this capability when the external wrench is unknown. Two criteria are reformulated using the L2L2 distance function, and further transformed into two nonlinear optimization problems. The differentiability of the objective functions and choice of initial conditions for global optimization are discussed. Keeping all the merits, that the criteria are applicable to grasps of 3-D objects with any contact types, and that the friction cones are not linearized, this work endows them with several new virtues: (a) Their formulation and computation are unified for both force-closure and non-force-closure grasps; (b) They are independent of the choice of coordinate frame and unit; (c) The object geometry is taken into account; (d) The computational efficiency is even higher than some methods by linearizing the friction cones.
