Utility Optimization Using Game-Based SG-DR Algorithm in V2G

This paper presents a utility optimization algorithm on Vehicle-to-Grid (V2G) aiming to maximize game based utility functions for Electric Vehicles (EVs) and the Aggregator. We first model this V2G as a "virtual" energy trading process by formulating a one-leader (aggregator) N-follower (EVs) Stackelberg Game (SG). In this game, both EVs and the aggregator try to maximize their own utility functions until they converge to a Game Equilibrium (GE). Moreover, we extend our game on a time duration by joining the Demand Response (DR) and introducing charging (G2V) to further optimizing the utilities. During this process, previous inputs and constraints of the game are constantly being updated, resulting the changes of GE. We formulate this combined SG-DR algorithm and run on a parking-lot which supports bi-directional energy and information exchange. Results show that our algorithm can achieve up to 50% of EV's utility increases than the pure SG.