Decentralized coordination of a building manager and an electric vehicle aggregator

The ability to control commercial buildings and electric vehicles (EVs) is a promising source of demand flexibility. In some cases, buildings and EVs share common infrastructure (e.g. a transformer) or interact with each other to accomplish a goal (e.g. reduce local peak demand). In such cases, the building and EV demand scheduling problems are effectively a single demand scheduling problem. Ideally, it would be solved as a single optimization problem. However, doing so might not be possible due to a number of concerns (e.g. data privacy). This paper proposes the use of a mixed-integer adaptation of the Dantzig-Wolfe decomposition to solve the building-EV demand scheduling problem in a decentralized fashion. The effectiveness of the proposed methodology is demonstrated in three case studies where the building and EV problems are coupled by either: i) demand limits, ii) a peak demand charge, or an iii) itemized billing tariff. Results show that the optimal solution can be reached while sharing a minimal amount of information. Furthermore, we show that the proposed methodology is scalable.