Neural network based adaptive stability control scheme for teleoperation under asymmetric time delays

In this paper, a novel neural network based control architecture is applied to the teleoperation system with asymmetric time delays. In the proposed method, two augmented error reference signals have been introduced to minimize the negative effects of time delays when interacting with slave environment. Generally speaking, the teleoperation system are subject to different types of uncertainties and unmodeled dynamics. In the proposed controller, the neural network estimates the nonlinear terms of the system and then the linearized system can be obtained. Using the concept of “adaptive estimation”, the unmodeled dynamic uncertainties are estimated with adaptive robust term to enhance the robustness of the controller. By the Lyapunov stability theory, we present the asymptotically stability condition of the closed-loop system which guarantees the uniformly ultimately bound of the neural network weights. Finally, experiments are simulated to validate the performance of the control method.