Optimizing the Uncertain Inverse Dynamics Bounds of the Franka Emika Panda Robot

In this paper, we propose an approach for the optimization of the uncertain inverse dynamics bounds of the Franka Emika Panda robot i.e., a robot manipulator with seven degrees of freedom. Our strategy is based on the use of interval arithmetic computations within a recursive Newton-Euler algorithm for inverse dynamics. By optimizing the bounds of dynamics parameters, we shrink the intervals describing the sets of torques of the residual dynamics found in experiments from nominal model knowledge. The specific uncertainty bounds of the Franka Emika Panda robot are disclosed. Experimental tests validate the results of the numerical optimization, demonstrate the applicability of the approach, and evaluate the influence of dynamics parameters bounds on robot joint torque sets.