Intelligent control for coordinating distributed energy storage

Stanford researchers have developed an architecture and control scheme for the coordination of distributed energy resources (DER), such as solar and storage, to minimize operation cost, enhance network reliability, and provide DER aggregation. The availability of electric load and generation data in electric grids varies both spatially and temporally due to: the random nature of loads and renewable generation; distributed locations of the loads and DER; and communication delays. Stanford researchers have found that control schemes that utilize local information (within the firmware of the storage unit, or in a home automation appliance) and delayed information to a global controller (e.g., collected from smart meter readings) perform almost as well as a perfect forecast global controller, which has all the load data (hence is not implementable).

Stanford's network architecture and control scheme simultaneously optimizes energy arbitrage and reduces energy cost by distributing control between a global controller (GC) and local controllers (LCs), shifting more responsibility for control decisions to the local controllers. This provides a win-win relationship between the utility companies, who may provide the global coordination, and the DER providers who may provide the local control.

Figure 1 - Smart grid with new technologies in the power transmission system

Figure 2 - Data flow in a distribution system. N customers with net loads S and local controllers LC. The local controllers communicate their data to a global controller GC, subject to a delay. The GC sends back control signals to the LCs based on the delayed information and its knowledge of the distribution network.