A gymnasium-compatible framework to create reinforcement learning (RL) environment for solving the optimal power flow (OPF) problem. Contains five OPF benchmark environments for comparable research.
Currently, when calculating the baseline OPF with pandapower, the self.net object is set to the optimal state, overwriting the previous state. This complicates comparisons of the two states because always only one is available. Also, such side effects may result in strange bugs.
Approach: Maybe store the optimized net in a different place.
Currently, when calculating the baseline OPF with pandapower, the
self.netobject is set to the optimal state, overwriting the previous state. This complicates comparisons of the two states because always only one is available. Also, such side effects may result in strange bugs.Approach: Maybe store the optimized net in a different place.