Apparatus and method for model predictive control (MPC) of a nonlinear process

An apparatus comprising:

• at least one memory configured to store:
• a process model that models a nonlinear process to be controlled, the nonlinear process comprising industrial production or processing equipment;
• an objective function that defines how the nonlinear process is controlled;
• and
• a control model that defines a dynamic feasible region associated with a controlled variable, the controlled variable associated with a physical characteristic of the process;
• and
• at least one processor configured to solve a control problem that includes the process model, the control model, and the objective function and to control the nonlinear process based on the solved control problem, wherein the at least one processor is configured to control the nonlinear process by generating and outputting a control signal to alter one or more manipulated variables associated with the process to thereby change the controlled variable;
• wherein the dynamic feasible region defined by the control model comprises a funnel region, the funnel region comprising a larger first portion and a smaller second portion following the first portion;
• and
• wherein greater flexibility is provided in meeting one or more control objectives for controlling the controlled variable in the first portion of the funnel region than in the second portion of the funnel region.

A tangible computer readable medium embodying a computer program, the computer program comprising:

• computer readable program code for solving a control problem that includes a process model, a control model, and an objective function, the nonlinear process implemented using industrial production or processing equipment;
• and
• computer readable program code for controlling the nonlinear process based on the solved control problem, wherein the computer readable program code for controlling the nonlinear process comprises computer readable program code for generating and outputting a control signal to alter one or more manipulated variables associated with the process to thereby change a controlled variable;
• wherein the process model models the nonlinear process to be controlled;
• wherein the objective function defines how the nonlinear process is controlled;
• wherein the control model defines a dynamic feasible region associated with the controlled variable, the controlled variable associated with a physical characteristic of the process;
• wherein the dynamic feasible region defined by the control model comprises a funnel region, the funnel region comprising a larger first portion and a smaller second portion following the first portion;
• and
• wherein greater flexibility is provided in meeting one or more control objectives for controlling the controlled variable in the first portion of the funnel region than in the second portion of the funnel region.