Real-time optimization modifier adaptation approach using quadratic approximation of the plant-model mismatch function

Abstract

The present work proposes a real-time optimization modifier adaptation scheme (RTO MA) in which the modifiers are obtained through a quadratic approximation of the plant-model mismatch function. The work also highlights the links between the model used in the trust-region framework, the model used in bias, gradient, and Hessian-based RTO MA schemes, and the model used in function-based RTO MA. In line with this, general conditions are established for the adapted nominal model for use on the basic trust-region algorithm. The proposed modifiers are then incorporated into an iterative algorithm of the real-time modifier adaptation scheme complemented with a trust-region framework. The effectiveness of the approach is demonstrated in simulation for the optimization of a continuous stirred tank reactor.