Vulnerability in dynamically driven oscillatory networks and power grids.

摘要: Vulnerability of networks has so far been quantified mainly for structural properties. In driven systems, however, vulnerability intrinsically relies on the collective response dynamics. As shown recently, dynamic patterns emerging in oscillator and AC power grid models are highly heterogeneous nontrivial, depending jointly driving frequency, interaction topology network, node or nodes driven. Identifying which most susceptible to may thus make system as a whole vulnerable external input signals, remains challenge. Here, we propose an easy-to-compute Dynamic Index (DVI) identifying those that exhibit largest amplitude responses signals with given spectra vulnerable. The DVI is based linear theory, such generic, enables robust predictions. It shows potential wide range applications across dynamically networks, instance, grids by fluctuating inputs from renewable energy sources output consumers.