Robust control using neural networks and input-output feedback linearization

This paper addresses the problem of stability robustness of minimum-phase nonlinear dynamical systems modeled with feedforward neural networks with bounded parametric uncertainties under IOF linearization. By means of an affine description of the feedforward neural network model which takes into account the parametric uncertainties, the Input-Output Feedback (IOF) linearization is performed and the uncertainty description is derived. Then a LMI-based robust controller is designed by means of an optimization procedure. A key step in this procedure is the derivation of a polytopic boundary for the state-space matrices of the IOF linearized system based on the estimated parameters of the neural network and bounds on their uncertainty.