Distributed Vehicle-to-Grid (V2G) Coordination: An Efficient and Cost-Effective Framework Based on Multilevel Game Theory

Due to the multiple attributes of transportation and energy storage device, electric vehicles (EVs) are expected to play an important role by vehicle-to-grid (V2G) interaction in future power system to cope with the high penetration of intermittent renewables. In order to improve the control efficiency while maintaining the cost-effectiveness of massive EV fleet scheduling, a hierarchical framework is proposed in this work, utilizing game theory to carry out economic incentives and control at three levels. The mean-field game and leader-follower Stackelberg game are combined in the lower and middle levels to model the competitive coordination among virtual power plants (VPPs) that aggregate the distributed EVs and to approximate the system state for extensive EV response. The evolutionary game is further employed in the upper level to examine the continuous configuration change with V2G adopted. Also, EV characteristic models of travelling features and battery degradation are established and simplified for efficient calculation. Case studies are implemented to verify the feasibility and performance. Results show that the required solution time can be reduced by distributed scheduling based on game method with similar and even better general utility achieved.