Wiener model based nonlinear predictive control

This paper addresses the problem of discrete-time nonlinear predictive cont rol of Wiener systems. Wiener-model-based nonlinear predictive control comb ines the advantages of lineal-model-based predictive control and gain sched uling while retaining a moderate level of computational complexity. A clear relation is shown between an iteration in the optimization of the nonlinea r control problem and the control problem of the underlying linear-model-ba sed method. This relation has a simple form of gain scheduling, thus the pr operties of the nonlinear control system can be analysed from the comprehen sible linear control aspect. Several disturbance rejection techniques ale p roposed and compared. The method was tested on a simulated model of a pH ne utralization process. The performance was excellent also in the case of a c onsiderable plant-to-model mismatch. The method can be applied as a first n ext step in cases where the performance of linear control is unsatisfactory owing to process nonlinearity.