Robustness of Aggregation in Networked Dynamical Systems

Luca Scardovi and Naomi Ehrich Leonard

In this paper we study robustness of aggregation in networks of coupled identical systems driven by possibly different external inputs. This property is guaranteed by providing a finite $L_2$ gain condition for the closed-loop system. We show, for a class of systems, how robustness depends on the connectivity of the underlying communication graph. Applications range from coordination problems where there are conflicting objectives to the study of aggregation phenomena where perturbations of the nominal systems must be taken into account. Both scenarios arise in networks of biological and engineered coordinating systems.