A CONSTRUCTIVE APPROACH FOR NONLINEAR-SYSTEM IDENTIFICATION USING MULTILAYER PERCEPTRONS

This paper combines a conventional method of multivariable system iden tification with a dynamic multi-layer perceptron (MLP) to achieve a co nstructive method of nonlinear system identification. The class of non linear systems is assumed to operate nominally around an equilibrium p oint in the neighborhood of which a linearized model exists to represe nt the system, although normal operation is not limited to the linear region. The result is an accurate discrete-time nonliner model, extend ed from a MIMO linear model, which captures the nonlinear behavior of the system.