The Worldwide Journal of Robotics Analysis, Forward of Print.
Robotic manipulation and locomotion typically entail nearly-simultaneous collisions—similar to heel and toe strikes throughout a foot step—with outcomes which are extraordinarily delicate to the order wherein impacts happen. Robotic simulators and state estimation generally lack the constancy and accuracy to foretell this ordering, and as a substitute decide one with a heuristic. This discrepancy degrades efficiency when model-based controllers and insurance policies realized in simulation are positioned on an actual robotic. We reconcile this subject with a set-valued rigid-body mannequin which generates a broad set of outcomes to simultaneous frictional impacts with any affect ordering. We first prolong Routh’s affect mannequin to a number of impacts by reformulating it as a differential inclusion (DI), and present that any answer will resolve all impacts in finite time. By contemplating time as a state, we embed this mannequin into one other DI which captures the continuous-time evolution of rigid-body dynamics, and assure existence of options. We lastly solid simulation of simultaneous impacts as a linear complementarity downside (LCP), and develop an algorithm for tight approximation of the post-impact velocity set with probabilistic ensures. We show our method on a number of examples drawn from manipulation and legged locomotion, and evaluate the predictions to different fashions of inflexible and compliant collisions.
发布者：Mathew Halm，转转请注明出处：https://robotalks.cn/set-valued-rigid-body-dynamics-for-simultaneous-inelastic-frictional-impacts/
