Adaptive neural-networks control for uncertain second-order fully actuated strict-feedback systems without over-parametrization

In this paper, the tracking control is addressed for second-order fully actuated strict-feedback systems with unknown dynamics. An adaptive neural-networks control law is constructed via command filtered backstepping control approach. Different from the existing traditional first-order state-space backstepping approach, the proposed control law can avoid "explosion of complexity", and does not need to convert the second-order systems into first-order form. Besides, the neural networks are utilized to identify unknown dynamics in the systems with only two tuning parameters. Based on Lyapunove stability theory, it is proven that the closed-loop system under the proposed adaptive neural-networks control law is stable in the sense of the uniformly ultimately bounded.