Optimization in Model Based Control

Abstract

This paper examines the design of controllers with emphasis on their ability to handle state and control constraints and nonlinearity. A major motivation is the presence of hard constraints in most applications, not least in the petro-chemical industry where steady state optimization forces the operating point to lie on or near the boundary of the feasible set. Efficient handling of constraints requires nonlinear controllers even if the system being controlled is linear and design of nonlinear systems involves optimization. The interaction between nonlinear control and optimization is explored, particular attention being given to model predictive control of constrained systems, linear and nonlinear.