Electrical vehicle grid integration for demand response in distribution networks using reinforcement learning

Most utilities across the world already have demand response (DR) programs in place to incentivise consumers reduce or shift their electricity consumption from peak periods off-peak hours usually financial incentives. With increasing electrification of vehicles, emerging technologies such as vehicle-to-grid (V2G) and vehicle-to-home (V2H) potential offer a broad range benefits services achieve more effective management demand. In this way, electric vehicles (EV) become distributed energy storage resources can conceivably, conjunction with other solutions, contribute DR provide additional capacity grid when needed. Here, an approach for V2G V2H using Reinforcement Learning (RL) is proposed. Q-learning, RL strategy based on reward mechanism, used make optimal decisions charge delay charging EV battery pack and/or dispatch stored back without compromising driving needs. Simulations are presented demonstrate how proposed effectively manage charging/discharging schedule smooth household load profile, minimise bills maximise revenue.