Hard constraints in control state variables of multivariable nonlinear processes resolved by elementary nonlinear decoupling

Abstract

A model based control strategy for nonlinear multivariable processes having hard constraints on both control and state variables, is derived based upon the concept of Elementary Nonlinear Decoupling (END). In END a property space is defined which has the same dimension as the control space and which permits inverse control. An optimization procedure is used to design a property transformation that gives the best solution in terms of performance, stability, robustness, etc. This control strategy is shown to be convenient when control and state variables are subject to hard constraints. When constraint limits are reached, new property variables are introduced which approximate the variables to be constrained and thereby achieving fast and accurate control on constraint boundaries.