A synthesis approach for output feedback robust model predictive control based-on input–output model

• New output feedback procedure for input–output model.
• Solution in the absence of state estimator.
• Recursive feasibility is guaranteed.
• The notion of quadratic boundedness is applied.

This paper proposes a synthesis approach to output feedback robust model predictive control for a system with both polytopic description and bounded disturbance. It is not required to estimate the system state since the input–output model is directly utilized. The procedure extends the main results of Kothare et al. (1996) in specifying the system model, feedback law and closed-loop stability. By on-line implementation of the controller, the output and input of the closed-loop system will converge to a neighborhood of the equilibrium point. An example of the continuous stirred tank reactor (CSTR) is given to illustrate the effectiveness of the controller.