Forecasting what Matters: Decision-Focused RL for Controlled EV Charging with Unknown Departure Times
Researchers have developed a decision-focused reinforcement learning (DF-RL) framework to improve electric vehicle (EV) charging control when departure times are unknown. This approach trains a forecaster and a charging policy controller end-to-end, allowing the forecaster to receive feedback on its impact on the controller's decisions. The DF-RL method demonstrated superior charging decisions compared to baselines, achieving up to a 14% improvement in total reward and a 55% reduction in unsupplied energy. AI
IMPACT This research could lead to more efficient and stable power grids by optimizing EV charging schedules.