The use of identified models in the control of a chaotic circuit

A number of control techniques require a model of the system to be controlled. There are many ways of estimating models from data obtained form the system. One way is to fit nonlinear polynomials with carefully chosen structure to the data available. The models obtained in such a way have a number of nice properties such a analytical tractability. This paper investigates the usefulness of such models in control problems in which a model is required. The numerical examples discussed in this paper use both monovariable and multivariable discrete models of Chua´s double scroll. The results of the paper suggest that the multivariable model is especially well suited for control problems