Bilateral Teleoperation System Control Based on Adaptive Neural Networks with Uncertainties and Time Delay

Abstract

A class of bilateral teleoperation systems with dynamics and kinematics uncertainties are investigated in this paper for synchronous tracking in the existence of communication time delay and unknown disturbances. Moreover, adaptive neural networks module is structured to approximate the uncertain dynamics, while the Jacobian matrix approximation technique is adopted to estimate the unknown kinematics, on the base of which the master-slave robots dynamics and kinematics without structurally known and linearizable parameters. By using the backstepping technique, the adaptive controllers are designed, and it is demonstrated via the Lyapunov method that the synchronous position and velocity tracking errors for master-slave robots eventually converge to a tiny neighbor near the origin. The final performance simulations verify the feasibility of the proposed controllers.