A Hierarchical Optimization Approach to Robot Teleoperation and Virtual Fixtures Rendering

Abstract

This paper presents a novel controller for bilateral teleoperation based on hierarchical constrained optimization techniques, that exploits motion constraints for the definition of virtual fixtures and haptic rendering. Our method allows the description of both rigid and compliant virtual fixtures, without the need of relative weight tuning as in previous approaches. Moreover, kinesthetic feedback due to each active constraint can be directly evaluated from the dual solution of the optimization algorithm, removing constraint penetration typical of penalty-based approaches and critical for telemanipulation where high accuracy is required. A transparency and stability analysis in case of communication delays is provided by relying on the classical two-port network formalism and linear control theory. An experimental evaluation of the approach for an object tracking task is carried out on a teleoperation system consisting of a 7-axis position-controlled slave robot with an eye-in-hand camera, and a 3 d.o.f. force-controlled master device.