Stochastic Model Predictive Control Using Indirect Feedback With a Modified Terminal Constraint
TL;DR
The main contribution is to explore the time‐varying definition of the constraints in the indirect feedback SMPC (IF‐SMPC) to improve the feasible domain and explore the time‐varying definition of the constraints in the indirect feedback SMPC to improve the feasible domain.
Abstract
This paper proposes a novel dynamic terminal set approach for the model predictive control (MPC) design. It concerns the class of linear systems subject to stochastic disturbances and builds upon an indirect feedback (IF) strategy. The main contribution is to explore the time‐varying definition of the constraints in the indirect feedback SMPC (IF‐SMPC) to improve the feasible domain. Using this feature, the closed‐loop performance of the indirect SMPC is not degraded, and the domain of attraction is enlarged, while the number of decision variables of the SMPC optimization problem is preserved. A constructive procedure is provided to obtain the sequence of time‐varying (evolving) terminal sets. Moreover, simulation case studies based on two benchmarks illustrate the notions and show the advantage of the new time‐varying terminal set approach for the IF‐SMPC.