Neural network implementation of nonlinear Receding-Horizon control

The Receding Horizon (RH) approach is an effective way to derive control al gorithms for nonlinear systems with stabilising properties also in the pres ence of state and control constraints. However, RH methods imply a heavy co mputational burden for on-line optimisation, therefore they are not suitabl e for the control of 'fast' systems, for example mechanical ones, which cal l for the use of short sampling periods. The aim of this paper is to show t hrough an experimental study how a Nonlinear RH (NRH) control law can be co mputed off-line, and subsequently approximated by means of a neural network , which is effectively used for the on-line implementation. The proposed de sign procedure is applied to synthesise a neural NRH controller for a seesa w equipment. The experimental results reported here demonstrate the feasibi lity of the method.