Environment Design Options¶
The OpfEnv base class provides various environment design options for representing the OPF problem as an RL environment. These options are available to all environments that inherit from the OpfEnv base class.
Overall, the RL environment design consists of the following main components:
Reward function definition
Observation space definition
Action space definition
Episode definition
Training and test data
Most environment design options are described in detail in Wolgast and Nieße - Learning the optimal power flow: Environment design matters.
TODO: Work in progress, more information will follow.