Optimal Bidding Strategy for Purchasing and Selling the Electricity by the EV Aggregator in the Energy Market, Based on Actor-Critic Algorithm

Great Advancements in Electric Vehicles (EVs) as well as their increasing utilization in smart grids has highlighted the key role of EV aggregators. As a mediator, EV aggregators are able to integrate EVs in power system structures, manage charging/discharging of EVs and decide whether or not purchase/sell the electricity on behalf of EV owners from/to electricity markets. On this basis, this entity can purchase the required energy of EVs with the lowest price during low-load hours and charge EV batteries. On the other side, it can sell the aggregated energy, resulted from discharging of EV batteries, with the highest possible price in the power market. As a consequence, this mediator can be economically beneficial both for the power grid and EV owners. Particularly, power companies will be able to shift load hours in the power grid. In this paper, through a bi-directional interaction with the competitive electricity market, the EV aggregator can be a retailer, purchase the energy from the power grid and provide the energy required by EVs. Additionally, it can be considered as a Generation Company, sell the aggregated energy in the competitive market and bring a considerable profit to EV owners. In this regard, an intelligent method is proposed with the aim of achieving the optimal bidding strategy, considering the day-ahead electricity market. The proposed approach is the actor-critic method, which is known as a class of Reinforcement Learning (RL) methods. Prominent advantages of the applied method include rapid learning capability, good convergence speed as well as high accuracy.