Optimization-based Reactive Force Control for Robot Grasping Tasks

It is proposed an optimization-based controller that reactively adapt the position of the end effectors in cooperative robot systems or fingers, in case of robotic hands. The proposed optimization-based controller uses the force tracking error for each robot, allowing to reactively correct the applied force, thus guaranteeing a stable grasp. Force information is used to determine the modification of the desired motion of the robots in a grasping task, so that ultimately the applied force to guarantee a stable object grasp is achieved. The novelty of our approach is that the proposed controller is obtained as the solution of a dynamic optimization problem, which is solved trough the standard gradient flow approach. Moreover, the method is free of in-depth models and its convergence properties are presented. Experimental results show that the proposed controller is effective, as far as an initial contact between the robots and the object is guaranteed.