Observer-based offset-free model predictive control for fractional-order systems

Model predictive control (MPC), with its unique features, has been widely used in the industry. Model accuracy is an important issue in predictive controllers. Fractional-order models can model real systems better than integer-order ones. Also, another significant issue in predictive controller design is ensuring zero steady-state tracking error, i.e., offset-free MPC. The Grunwald-Letnikov approach is used to design an offset-free MPC for fractional-order systems in this article. A two-step method for designing offset-free MPC is used so the predictive control problem is wholly separated from the observer design. The gain of the observer is obtained through an optimization problem to decrease the effects of disturbances. Finally, the effectiveness of the proposed method is shown via two numerical examples.