Adaptive neural network tracking control of snake-like robots via a deterministic learning approach

This paper proposes a new learning control method for snake-like robots to achieve trajectory tracking. Based on deterministic learning, an adaptive neural networks control algorithm is used to track the desired trajectory and approximate the unknown system dynamics of the snake-like robot. After that the learned knowledge from direct neural networks is stored as constant network weights. These weights improve the response speed and the accuracy of the system in repeating same or similar control tasks. By using the Lyapunov approach, the tracking error is proven to be uniformly ultimately bounded and converges to a residual set. Finally, simulation results are presented to illustrate the effectiveness of the proposed control scheme.