Elasticity-Aware Online Motion Optimization for Link-Elastic Manipulators

Abstract

The field of human-robot interaction is a typical application of elastic robots, as they reduce the risk of injuries and physical damage in case of a collision. Elasticities, however, also impose high demands on underlying joint controllers to guarantee minimal vibration during regular operation. Numerous control concepts assume a sufficiently high ability to control vibrations, by e.g., dedicated actuators or special kinematic structures. This work presents an online, optimization-based trajectory planning approach that concentrates on maximizing this ability for elastic manipulators without additional damping actuators or certain kinematic structures. The planning algorithm utilizes a modified quadratic objective function to incorporate the controllability of vibrations as a secondary goal. The effectiveness of the approach is demonstrated on a real 3-DOF, link-elastic robot for different set-points subject to disturbances. The results show that the approach successfully generates elasticity-aware motions and improves the vibration damping capabilities of the underlying controllers. Especially for critical configurations in which the controllers usually have little or no influence on the vibrations, vibration damping is improved or even made possible.