Real-Time Optimization via Modifier Adaptation Integrated with Model Predictive Control

A. MARCHETTI; P. LUPPI; M. BASUALDO

Milán

Congreso; 18th IFAC World Congress, 2011; 2011

IFAC

In order to deal with plant-model mismatch, real-time optimization schemes use someadaptation strategy based on measurements. The modifier-adaptation approach is to correct theconstraints and gradients in the optimization problem by adapting the values of bias modifiersexpressing the difference between the constraints and gradients of the plant and the model. Theapproach has the ability to converge to the plant optimum but does not guarantee feasibilityprior to convergence if the evaluated optimal inputs are applied directly to the plant. In thispaper, an approach for integrating modifier adaptation with model predictive control is presentedwhere both automation layers use the same decision variables. The control targets are includedas equality constraints in the real-time optimization problem, and are continuously enforced bythe model predictive controller. In order to apply modifier adaptation, new gradient modifiersare defined that correct the projected gradients in the tangent space of the equality constraints,rather than the full gradients. An illustrative case study shows the applicability of the approach.