“…If the literature of model-based predictive control is studied, the most conspicuous feature is that the choice of an adequate (sometimes nonlinear) model has a distinguished role in the control design procedure of these controllers. The solutions range from the application of impulse response and step response models , to the employment of neural network models , and fuzzy models , or recently to Hammerstein models , or ARMA and CARIMA models. , The main improvements in the models, yielded with time, consist in their nonlinear feature.…”