par Delalieux, Yves
Référence Lecture notes in control and information sciences, page (174-179)
Publication Publié, 1985
Article révisé par les pairs
Résumé : Some large scale systems are hardly optimized : their optimal operating point cannot be directly found because the system models are unknown or difficult to be approximated. In that case, real-time optimization can be achieved by choosing successive operating points in such a way that the system operates nearer its optimal operating point. These various points inform about the relationships between the inputs and the outputs variables. Hierarchical optimization algorithms are presented which allow optimizing simultaneously several criteria.