Decentralized economic dispatch of an isolated distributed generator network

A novel decentralized method for optimal load distribution in an isolated power system is proposed. In contrast to the traditional centralized method, the load economic dispatch is distributed to every smart distributed generator (DG) without the need for a monitoring host. Similar to the structure, mechanism and characteristics of biological communities, a smart DG node can communicate with adjacent nodes and operate collaboratively to complete the optimal operation of an isolated power system. The task is formulated as a decentralized optimization with a number of local constraints and is solved by a penalty function method. With the economic dispatch algorithm computed in all of the DG nodes in parallel, a new, fully distributed flat control network is established. The convergence property of the novel method is analysed theoretically. Simulation results on an illustrative system provide support for the validity of the proposed method.