Consensus Dynamics in Switching Networks for Distributing Power Packets

Power packet dispatching system is one of promising technologies for energy saving. To enhance this system, we propose an efficient algorithm for distributing power packets. As the first step, we consider the consensus dynamics in a network by assuming the state variable at each node as an amount of power. In addition, we assume that networks are switching, namely the topology of a network changes in time. These assumptions are effective for distributing power packets among power network systems in terms of decentralized control. We demonstrate numerical simulations for a small-scale network and discuss the results for an application to the power packet dispatching systems, such as variability of power storage at a node, control of power distribution over a network, and topology estimation from time series.