Switching adaptive control of a class of non-affine nonlinear systems

In this paper, a performance-oriented switching adaptive control based on adaptive neural networks control and model-free adaptive control is proposed for a class of non-affine discrete time nonlinear systems. The method is composed of adaptive neural networks control, model-free adaptive control and a switching mechanism. The adaptive neural networks control ensures the transit performance and the model-free adaptive control is used to improve the final tracking performance. Switching scheme orchestrating between the two adaptive control algorithms can guarantee both the stability of controlled system and the control performance simultaneously. Stability and convergence analysis are given for the closed-loop system as well. Finally, simulation examples is presented to illustrate the effectiveness of the presented methods.