Modeling and classification of non-linear systems using neural networks--I. Simulation

Abstract It is shown that multi-layer perceptron (MLP) neural networks can be used to model accurately sdof non-linear dynamical systems and to separate unambiguously linear and non-linear responses. The networks are also applied successfully to the problem of classifying the type of non-linearity present in the system. The data used is obtained from numerical simulation; a validation method for non-linear system simulation is presented based on higher order frequency response functions.