Discrete-Time Dynamic Output Feedback Integral Sliding Model Predictive Control

In this paper, a novel discrete-time dynamic output feedback integral sliding model predictive control is proposed for linear systems. A Luenberger observer is designed when the states are not completely measurable. A set of inequalities are presents to keep the practical states, observed states and observation errors within the same ellipse, then input and state constraints can be solved by linear matrix inequalities (LMI). A new sliding surface is proposed which is relevant to the system observer states and model predictive control law. It is also proved that this method can effectively guarantee the stability of the closed-loop system. Finally, the validity of proposed method is verified by simulation.