NEURAL AND ADAPTIVE CONTROLLERS FOR A NONMINIMUM-PHASE VARYING TIME-DELAY SYSTEM

In this paper we study a non-minimum phase discrete time system with v arying time-delay. We first propose several open loop control architec tures based on non-linear neural networks and study their ability to h andle the different difficulties of the control problem. All the metho ds are tested and compared to a baseline linear controller, on a simul ated river system. This plant is submitted to perturbations correspond ing to water withdrawals and lateral inflows. The above architectures are not able to cope with such perturbations. We then propose a model combining a feed-forward neural network based learning controller and a feedback adaptive controller. The performances of this model are com pared to a similar architecture containing linear feed-forward and fee dback controllers. (C) 1997 Elsevier Science Limited.