Adaptive control of a class of nonlinear discrete-time systems using hybrid neural networks

In this paper, an indirect adaptive controller based on hybrid neural networks, which are composed of two-layered neural networks and radial basis function (RBF) neural networks, is derived for controlling a class of unknown nonlinear discrete-time systems. A hybrid-neural-network-based estimator is used to characterize the input-output behavior of the unknown systems. An adaptation law which adjusts the connection weights of the neural network is used to minimize the error signal which is difference between the actual response and that of the neural network. The indirect adaptive control law is generated on-line using another hybrid neural network related to the estimator, so that the plant results in a bounded tracking error with respect to a desired reference signal. It is proved that the control objective is achieved by the closed-loop system and that the system remains closed-loop stability. The effectiveness of the proposed control scheme is also demonstrated by a simulation example.