Inclusion of Actuator Saturation as Complementarity Constraints in Integrated Design and Control

Abstract

The design of a process can significantly affect its ability to be satisfactorily controlled, which motivates the need to design the process and its associated control system simultaneously. However, since all physical systems have constraints, it is necessary to include these within the model as well. A bilinear formulation for rigorous saturation handling is able to correctly describe the logical conditions which accompany saturation of the manipulated variable and is suitable for incorporation within a simultaneous optimization framework. This paper explores a number of issues. First, a bilinear formulation is used to examine the effect of including anti-reset windup within an integrated design. Second, the performance of a solver specifically tailored to solve optimization problems with complementarity constraints is compared to other nonlinear solvers for the solution of a simple nonlinear case study. Finally, its performance on a more complex nonlinear CSTR case study is discussed.