Robust object manipulation for prosthetic hand applications

We propose a robust control law for the manipulation of objects in light of disturbances that result from lack of a priori knowledge of the grasp scenario. An application of this control law is in cases where it is impractical to assume all properties of the object and external wrenches are known to the controller, notably in cases of object grasping and manipulation with prosthetic hands. The main contribution of the proposed approach is to guarantee semi-global asymptotic stability and exponential stability of the system. In addition, the approach lends itself to a systematic tuning method presented in the literature that improves transient performance without compromising stability of the system. Simulation results validate the effectiveness of the proposed approach.