Wave-based teleoperation with prediction

Abstract

Wave variables based on passivity and scattering theory provide a good tool for establishing bilateral teleoperation in the presence of a constant time delay. Recently these techniques have been extended to be used for a varying time delay, as in the case of Internet based teleoperation. Although stability is guaranteed for virtually any time delay, performance rapidly degrades for larger delays. In this paper a predictor derived from a modified Smith predictor along with a Kalman estimator and an energy regulator is used to enhance the performance of a wave-based teleoperator in the presence of a constant delay. Also the current wave transformation equations are extended to a more general case (better suited for systems with multiple degrees of freedom).

Robustly stable bilateral teleoperation under time-varying delays and data losses: an energy-bounding approach

“…Thus, the proposed EBA does not need to adjust 1 c , 2 c parameters according to time delays unlike other algorithms-e.g., characteristic impedance b in wave variable approach [10][11][12][13][14][15][16] and dissipation gain d K in Lee and Spong's framework [21]. This advantage is demonstrated in Section 4 through a series of experiments with severe timevarying delays and packet losses.…”

“…2-4, 6 and 7 are observed before and after contact with a wall due to the added dissipation term. In order to take into account those problems, additional approaches such as a wave predictor [14,15], drift compensation [15], or reference energy following [18], have been proposed. The experimental results of the EBA, however, show none of the aforementioned drawbacks of other approaches.…”

“…Moreover, the proposed EBAs do not transform any variables, as in Refs. [9][10][11][12][13][14][15][16]. It only restricts the excessive energy to the energy limit that can be consumed by the energy dissipation capability in the system.…”

“…However, a steady-state error was observed due to the variable time delay between the master and the slave, referred to as position drift. To cope with this problem, additional control methods such as virtual time delay, drift compensation, wave prediction, and explicit position data transmission through additional paths have been proposed [12][13][14][15][16]. A time domain passivity algorithm, initially proposed for stable haptic interactions [7] and was later extended to telepresence systems with constant time delays by Artigas et al [17][18][19], ensures stability based on a passivity observer and a passivity controller.…”

“…Moreover, passivity can ensure stability without requiring the exact knowledge of operator and environmental models. For this reason, it has been used extensively for the analysis of coupled stability problems in robotics [3,4], haptic [5][6][7][8] and teleoperation systems [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24].…”