Real-time optimization of energy systems in sugar and ethanol facilities

Abstract Real-time optimization (RTO) is used to operate industrial processes close to their minimum cost or maximum profit conditions. Model-based RTO needs an adaptation step to update the model under plant/model mismatch. This work proposes a novel RTO approach based on the modifier adaptation strategy, aiming to reduce the dimension of the gradient correction problem. The approach was tested in a model of a heat and power system of a sugar/ethanol facility. It could be also applied to other kind of processes. The results showed a noticeable performance improvement when compared to adaptation strategies without gradient correction.