An alternative real-time optimization algorithm

FERNAN SERRALUNGA; MIGUEL C MUSSATI; AGUIRRE PIO

Londres

Congreso; Proceedings of the 22nd European Symposium on Computer Aided Process Engineering; 2012

Performance of industrial processes can be continually improved through real-time optimization (RTO). Model-based RTO approaches always deal with parametric and structural plant/model mismatch, and adaptation strategies have been developed to overcome this problem. This work proposes an alternative algorithm based on the modifier adaptation approach, which simplifies the gradient estimation problem in systems with a large number of variables. The algorithm is tested in a heat and power system model, which includes a steam network with electric power generation.