Contrôle optimal en boucle fermée d’instabilités hydrodynamiques par identification de modèles réduits
AURÉLIEN HERVÉ, PhD Thesis
system identification, model reduction, flow control, fluid mechanics, non linear control, POD, ARX, ARMAX, NARMAX
This work aims at defining a low order modeling framework, in order to perform flow control. Two different flow configurations are introduced, that respectively represent a typical case of an amplifier flow and an oscillator flow. For both flows, system identification techniques along with a physical study of the flow are used in order to define a low order model that is suited for control purposes. An efficient closed loop control system is then built, and used to suppress the flow unsteadiness.