跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2604.01830 · v2

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

首次提交 2026-04-02 17:45
版本更新 2026-11-17 06:28
雷达观测 2026-11-18 19:43
arXiv ↗PDF ↗

摘要

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 computationally expensive. We propose a physics-informed Reinforcement Learning framework that combines semi-Markov control with a Gibbs prior, that encodes the system's physics, over the action space. The decision is only taken when the grid enters a hazardous regime, while a graph neural network surrogate predicts the post action overload risk of feasible topology actions. These predictions are used to construct a physics-informed Gibbs prior that both selects a small state-dependent candidate set and reweights policy logits before action selection. In this way, our method reduces exploration difficulty and online simulation cost while preserving the flexibility of a learned policy. We evaluate the approach in three realistic benchmark environments of increasing difficulty. Across all settings, the proposed method achieves a strong balance between control quality and computational efficiency: it matches oracle-level performance while being approximately 6×6\times faster on the first benchmark, reaches 94.6%94.6\% of oracle reward with roughly 200×200\times lower decision time on the second one, and on the most challenging benchmark improves over a PPO baseline by up to 255%255\% in reward and 284%284\% in survived steps while remaining about 2.5×2.5\times faster than a strong specialized engineering baseline. These results show that our method provides an effective mechanism for topology control in power grids.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v2,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v2 · 2026-11-17 06:28 · 最新观测版本

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

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 computationally expensive. We propose a physics-informed Reinforcement Learning framework that combines semi-Markov control with a Gibbs prior, that encodes the system's physics, over the action space. The decision is only taken when the grid enters a hazardous regime, while a graph neural network surrogate predicts the post action overload risk of feasible topology actions. These predictions are used to construct a physics-informed Gibbs prior that both selects a small state-dependent candidate set and reweights policy logits before action selection. In this way, our method reduces exploration difficulty and online simulation cost while preserving the flexibility of a learned policy. We evaluate the approach in three realistic benchmark environments of increasing difficulty. Across all settings, the proposed method achieves a strong balance between control quality and computational efficiency: it matches oracle-level performance while being approximately 6×6\times faster on the first benchmark, reaches 94.6%94.6\% of oracle reward with roughly 200×200\times lower decision time on the second one, and on the most challenging benchmark improves over a PPO baseline by up to 255%255\% in reward and 284%284\% in survived steps while remaining about 2.5×2.5\times faster than a strong specialized engineering baseline. These results show that our method provides an effective mechanism for topology control in power grids.

查看此版本解读 →查看此版本原文 ↗