Distributed Continuous-Time Algorithm for Economic Dispatch Problem Over Switching Communication Topology

This brief mainly studies the economic dispatch problem (EDP) in smart grid. The purpose of this problem is to minimize the total power generation cost under the conditions of balance between supply and demand and power generation capacity constraints. A distributed continuous-time solver is proposed, which only requires the sign of relative dual variable information between neighbors, which ensures the privacy of the generators. In addition, the algorithm does not require initialization and allows online changes of network topology, load, etc. It is theoretically proved that the algorithm achieve the optimal scheduling of smart grid over undirected switching graphs under certain conditions. In the presence of gradient measurement noise, the variation form of the algorithm which converges in the mean square sense is given. Finally, the simulation example illustrates the validness of the algorithm.