Neural Network Modelling and Control of a Plant Exhibiting the Jump Phenomena

Abstract

This paper focuses primarily on the modelling and control of nonlinear systems that exhibit gain discontinuities in their frequency plots. Structures and learning algorithms for neural network based nonlinear modelling are introduced and applied to the modelling and k-step ahead prediction of an example system. A nonlinear Internal Model Controller (IMC) is developed, based on the ability of the feedforward neural network to form nonlinear forward and inverse models. The results of simulation studies are given in each case.