Discrete-time multi-model control for cooperative teleoperation under time delay

Abstract

While a conventional bilateral teleoperation system involves only one pair of master/slave robots, cooperative telerobotic systems can consist of multiple pairs of robotic manipulators. Due to dynamic interaction among slave manipulators as well as communication latency, the control of such systems can be particularly challenging. This paper presents a multimodel discrete-time controller for teleoperation in cooperative environments subject to a known constant communication delay. Discrete-time state-space models that explicitly incorporate signal delays are developed for free motion/soft contact and rigid contact phases of teleoperation. Mode-based linear quadratic Gaussian (LQG) controllers are proposed that can deliver a stable transparent response for each phase of operation. Switching among these controllers occur according to the identified phase of contact. The robustness of the controllers with respect to parametric uncertainty is examined via the Nyquist analysis. Simulation results demonstrate the effectiveness of the proposed approach