Fuzzy Neural-Network-based Output Tracking Control for Nonlinear Systems with Unknown Dynamics

In this paper investigates the control problem for a class of nonlinear systems by utilizing fuzzy neural network (FNN). Fuzzy logic systems (FLSs) have proved to be an efficient technique in approximating the unknown continuous functions. However, its approximation accuracy highly rely on suitable membership functions and fuzzy rules, which are very difficult to choose. The introduction of neural networks (NNs) can help to adjust the membership functions and fuzzy rules by learning mechanisms. Therefore, FNN is utilized to approximate the unknown dynamics of the studied nonlinear systems. It should be noticed that the weights of the FNN are updated online instead using adaptive technique. In order to obtain the derivative estimations of error signals, which are used to reconstruct the unknown approximation errors, a differentiator is employed. Then, Lyapunov stability theory combined with back-stepping technique are used to design controller and guarantee that the system outputs track the target signals within a very small error range. Finally, simulation and comparison studies have been demonstrated to verify the effectiveness and merits of the proposed method.