Bi-level Distributed Optimal Dispatch of Micro Gird Clusters Based on Mutual Communication

Micro Gird Cluster (MGC), connecting to the medium or low voltage distribution network, usually contains certain number of Micro Grids, distributed generators, energy storages and loads. Considering coordinated operation in MGC, this paper proposes a hierarchical control architecture- the upper layer operated in distributed mode among multiple MGs and the lower layer were controlled by each micro grid center. This hierarchical architecture not only shows the excellent features of distributed control, but also integrates the advantages of centralized control. According to this hierarchical architecture, bi-level distributed optimal dispatch of MGC is analyzed. With the aim of minimizing MGC total costs in upper layer and minimizing each Micro-Grid costs in lower layer, bi-level optimal mathematic model take global and local aims into account synthetically. The distributed primal-dual sub-gradient algorithm is designed to solve the upper layer optimal model, alternating gradient iteration algorithm for the lower optimization model. Simulation cases show that bi-level distributed optimal mathematic model successfully fulfill the aim of coordinated operation in MGC, and distributed primal-dual sub-gradient algorithm is an effective solution to this optimal model.