Economic model predictive control for nonlinear processes incorporating actuator magnitude and rate of change constraints

Abstract

In this work, we consider nonlinear systems and develop a formulation of Lyapunov-based economic model predictive control (LEMPC) that incorporates actuator magnitude and rate of change constraints on the calculated control actions. Specifically, we formulate LEMPC including maximum and minimum allowable values for the control actions as well as bounds on the differences between the control actions calculated by the LEMPC and the control actions that would be requested if a stabilizing Lyapunov-based controller were used to control the system. This latter constraint is proven to bound the rate of change of the actual process inputs. It is also proven that an LEMPC formulation incorporating such constraints is feasible and ensures closed-loop stability of the process for a sufficiently small controller sampling period. A chemical process example demonstrates that EMPC with input rate of change constraints can improve process economic performance compared with steady-state operation while limiting actuation energy expended.