Bilateral motion prediction and control for teleoperation under long time-varying delays

Bilateral controller design for the teleoperation system is studied in this paper based on a motion prediction approach. To compensate the known long time-varying delays, novel predictors are presented to reconstruct the positions and velocities of robots on both sides through using the delayed measurements. The proposed predictors consist of several sub-predictors in a cascade structure, each of which is to predict the states of the previous one. The estimations of the actual states can be obtained from the last sub-predictor. New prediction horizons of each sub-observers are designed to cope with the time-varying fractions of the time delay. Then through applying the predicted results, bilateral predictive controller is designed for the teleoperation. The errors of both position tracking and prediction can converge into the bounded regions under several sufficient conditions of the control gains, which are obtained by using the Lyapunov-Krasovskii approach. The effective capacity of the presented method can be verified through comparative simulations.