Feedback control of flow resonances using balanced reduced-order models

This paper investigates the use of balanced reduced-order models for the feedback control of flow resonances. Specifically, the Eigensystem Realization Algorithm is used to find balanced reduced-order models of the linear dynamics of such flow resonances. The method is applied first to a computational problem in direct numerical simulations of cavity resonances, and then to a lab-scale experiment of combustion oscillations. Although the resulting reduced-order models both have fewer than 10 degrees of freedom, the feedback controllers that are based on them perform very well, with closed-loop stability achieved over a wide range of operating conditions.