Multiagent Charging of Electric Vehicles Respecting Distribution Transformer Loading and Voltage Limits

In this paper, a market based multiagent control mechanism that incorporates distribution transformer and voltage constraints for the charging of a fleet of electric vehicles (EVs) is presented. The algorithm solves a utility maximization problem in a distributed way, assigning most charging power to the EVs with the highest need for energy. The algorithm does not rely on an iterative exchange of messages, but finds the optimal solution after the exchange of just one single message. A substation agent is responsible for guaranteeing a safe network operation. It uses the remaining capacity of each of the EV chargers for reactive voltage control. The performance of the algorithm is evaluated on an existing three-phase four-wire distribution grid. Simulation results show that the fleet of EVs can be charged at a minimal increase of costs, without jeopardizing the network.