Distributed optimal batteries charging control for electric vehicles fleet
Large number of electrical vehicles can take significant role in energy consumption of buildings and homes. Connection to the grid simultaneously offer additional possibilities of electrical energy storing in price-optimal balancing of energy buying and consumption. The focus in the thesis is put to the optimization of generic number of electrical vehicles with variable time instants and duration of grid connectivity while assuring the required battery state of charge when disconnecting from the grid. The total charging price of all the connected electrical vehicles is minimized while taking into account the rest of building consumption profile plan, volatile electricity prices and battery wear-off cost due to limited number of charging and discharging cycles. The battery charging control is performed by a distributed game theory approach in model predictive control with a central supervisor. The supervisor manages the overall battery energy exchange and provides the scarce information to distributed local optimization problems for different electrical vehicles and batteries characteristics. Hourly prices and available energy are provided by the outside system, whether the utility grid or higher-level control system.
Keywords: Electrical vehicles fleet, battery management system, distributed model predictive control, game thoery