| Definition | The Building Energy Storage Simulation is a gymnasium-based environment for optimizing battery and solar energy management to minimize electricity costs using RL. |
| No of environments | 1 |
The Building Energy Storage Simulation is a gymnasium-based environment designed for reinforcement learning experiments in energy management systems. The environment models a building equipped with solar panels and a battery energy storage system, connected to a power grid with dynamic electricity pricing. The task involves controlling the battery’s charging and discharging actions to minimize overall electricity costs by optimizing energy usage, especially during peak pricing periods or when there is excess solar generation. Importantly, the environment restricts energy export to the grid, meaning surplus solar energy is lost if not stored.
This environment simplifies the complexity of energy management simulations by providing an easy-to-use interface, inspired by the CityLearn environment but focused on more accessible implementation. With customizable forecasting horizons for electric load, solar generation, and electricity prices, the environment challenges algorithms to balance energy consumption and storage strategies efficiently. The framework supports various solution approaches, including deep reinforcement learning, optimal control, and model predictive control.
| Example observations | State of Charge, Forecast Electric Load, Forecast Solar Generation, Electricity Price |
| Example actions | Charge |
| Example reward | rt=−1∗electricity_consumed∗electricity_price |