In this paper, a two-layer network and distributed control method is proposed, where there is a top-layer communication network over a bottom-layer microgrid. The communication network consists of two subgraphs, in which the first is composed of all agents, while the second is only composed of controllable agents. The distributed control laws derived from the first subgraph guarantee the supply–demand balance, while further control laws from the second subgraph reassign the outputs of controllable distributed generators, which ensure active and reactive power are dispatched optimally. However, for reducing the number of edges in the second subgraph, generally a simpler graph instead of a fully connected graph is adopted. In this case, a near-optimal dispatch of active and reactive power can be obtained gradually, only if controllable agents on the second subgraph calculate set points iteratively according to our proposition. Finally, the method is evaluated over seven cases via simulation. The results show that the system performs as desired, even if environmental conditions and load demand fluctuate significantly. In summary, the method can rapidly respond to fluctuations resulting in optimal power sharing.
展开▼
