Neural networks based NARX models in nonlinear adaptive control

The paper discusses the applicability of approximate NARX models of nonlinear dynamic systems. The models are obtained by a new version of Fourier analysis based neural network also described in the paper. This is a reformulation of a method, already presented, in a recursive manner, i.e., adapted to account for incoming data online. The method allows us to obtain an approximate model of the nonlinear system. The estimation of the influence of the modelling error on the discrepancy between the model output and real system output is given. The possible applications of this approach to the design of BIBO stable closed-loop control are proposed.