research18h
Forecasting what Matters: Decision-Focused RL for Controlled EV Charging with Unknown Departure Times
Researchers propose a decision-focused reinforcement learning approach for controlling electric vehicle charging when departure times are unknown. The method learns effective charging policies from historical data despite missing key features. This approach aims to reduce peak demand and grid instability caused by EV adoption.
Key takeaways
- Decision-focused RL handles unknown departure times in EV charging.
- Method learns from historical data to optimize charging policies.
- Aims to mitigate peak demand and grid instability from EV growth.