Interpretable Policy Distillation for Power Grid Topology Control

2026-05-30 · Aleksandra Dmitruka, Karlis Freivalds arxiv

Deep reinforcement learning (RL) offers a promising route to real-time power grid operation, yet large neural policies are costly to evaluate, hard to deploy on constrained hardware, and opaque to operators. We ask whether a Proximal Policy Optimization (PPO) agent for grid topology control can be compressed into compact tree-based surrogates without losing operational performance. A PPO teacher is trained on Grid2Op's standard 14-bus environment with a stability-oriented reward, using stress-focused data collection on critical, high-loading states. The policy is then distilled into a decision tree and a random forest. Across held-out validation episodes, both surrogates exceed the teacher in mean reward and survival length at a fraction of the inference cost. The decision tree shows high exact-action agreement with the PPO argmax and near-complete agreement within its top-ranked actions, while remaining small enough to be inspected directly. Feature-importance analysis reveals a representational shift: the PPO policy relies mainly on line-loading signals, while the distilled tree is driven primarily by bus-topology variables. These results suggest that stress-focused distillation can convert a black-box neural controller into a lightweight, auditable rule-like surrogate suited for real-time deployment, while also surfacing risks tied to deterministic actions and topology-specific generalization.

📄 PDF Abstract BibTeX arXiv:2606.00561

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Multi-Objective Reinforcement Learning for Power Grid Topology Control

2025-01-27 · Thomas Lautenbacher, Ali Rajaei, Davide Barbieri, Jan Viebahn 외

Transmission grid congestion increases as the electrification of various sectors requires transmitting more power. Topology control, through substation reconfiguration, can reduce congestion but its potential remains und…

Multi-Objective Reinforcement Learning reinforcement-learning Reinforcement Learning

Physics Informed Reinforcement Learning with Gibbs Priors for Topology Control in Power Grids

2026-04-02 · Pantelis Dogoulis, Maxime Cordy arxiv

Topology control for power grid operation is a challenging sequential decision making problem because the action space grows combinatorially with the size of the grid and action evaluation through simulation is computati…

Computational Efficiency Reinforcement Learning Graph Neural Network Decision Making

California Test System (CATS): A Geographically Accurate Test System based on the California Grid

2022-10-09 · Sofia Taylor, Aditya Rangarajan, Noah Rhodes, Jonathan Snodgrass 외

This paper presents the California Test System (CATS), a synthetic transmission grid in California that can be used by the public for power systems policy research without revealing any critical energy information. The p…

Fully Decentralized Reinforcement Learning-based Control of Photovoltaics in Distribution Grids for Joint Provision of Real and Reactive Power

2020-08-03 · Rayan El Helou, Dileep Kalathil, Le Xie

In this paper, we introduce a new framework to address the problem of voltage regulation in unbalanced distribution grids with deep photovoltaic penetration. In this framework, both real and reactive power setpoints are …

reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)

SOM-VQ: Topology-Aware Tokenization for Interactive Generative Models

2026-02-24 · Alessandro Londei, Denise Lanzieri, Matteo Benati arxiv

Vector-quantized representations enable powerful discrete generative models but lack semantic structure in token space, limiting interpretable human control. We introduce SOM-VQ, a tokenization method that combines vecto…