Adaptive non-linear predictive control of a simulated flotation cell

This work illustrates the non-linear modelling and control of a flotation cell using a predictive strategy. A non-linear predictive model, based on local linear models, is used to represent the predictions of the system output and to calculate the control signal. This model is directly identified, using the available input-output data, and then used to calculate the control signal. Simulated results show not only the advantages of using simple models to approximate complex phenomenological models, but also the performance of the proposed predictive control approach.