Adaptive controller for marginally stable nonlinear systems using neural networks

A method for the adaptive control of a marginally stable discrete-time nonlinear system of unknown structure is presented. The method is based on the model reference control technique in which the output from the nonlinear plant is made to track the output of a stable reference model which reflects the desired dynamics. Multilayered neural networks are used to approximate the plane dynamics and to implement the controller. Simulation studies show that the proposed scheme performs well under different working environments. Various design parameters that have significant influence on the performance of the overall system are discussed. The effect of the learning rates is highlighted.<>