Offset-free reference tracking for predictive controllers

We discuss the offset-free reference tracking problem for linear constrained systems in the presence of disturbances and plant-model mismatch. Introducing disturbance models, we show how to construct a controller/observer combination, such that zero offset is achieved. Contrary to other approaches, the plant model is augmented only by as many disturbance states as there are tracked variables, thus yielding the controller with minimal complexity according to the internal model principle.