Distributionally Robust Safety Filter for Learning-Based Control in Active Distribution Systems
This addresses safety issues for power system operators during AI-based control training, but it is incremental as it builds on existing safety filter methods with robustness enhancements.
The paper tackles the problem of operational constraint violations when deep reinforcement learning agents train in active distribution systems, by introducing a distributionally robust safety filter that significantly reduces violations while maintaining near-optimal solutions, as verified on IEEE 33-bus and 123-bus systems.
Operational constraint violations may occur when deep reinforcement learning (DRL) agents interact with real-world active distribution systems to learn their optimal policies during training. This letter presents a universal distributionally robust safety filter (DRSF) using which any DRL agent can reduce the constraint violations of distribution systems significantly during training while maintaining near-optimal solutions. The DRSF is formulated as a distributionally robust optimization problem with chance constraints of operational limits. This problem aims to compute near-optimal actions that are minimally modified from the optimal actions of DRL-based Volt/VAr control by leveraging the distribution system model, thereby providing constraint satisfaction guarantee with a probability level under the model uncertainty. The performance of the proposed DRSF is verified using the IEEE 33-bus and 123-bus systems.