Data-Assimilation-Assisted Reinforcement Learning for Power Grid Control under Load Uncertainty
Abstract
Reliable power grid control requires sequential decisions under transmission constraints, time-varying demand, renewable variability, and imperfect load information. We investigate how load information quality affects control in a customized transmission network simulator inspired by Grid2Op and based on a DC power flow model. The environment includes stochastic spatial loads, temporally correlated renewable-like generation, energy storage, line protection, generator dispatch, and line reconnection. An Ensemble Score Filter (EnSF) corrects noisy load information before it is passed to the controller. A heuristic study first isolates the effect of this correction under a fixed feedback rule. Proximal policy optimization (PPO) agents are then trained for sequential generator and reconnection control, frozen, and evaluated with Forward, EnSF-corrected, and Truth load information on paired stochastic scenarios. EnSF consistently lowers load estimation error. Under heuristic control, the correction extends survival and reduces warning and overload exposure. With frozen PPO policies, it increases cumulative return and keeps performance closer to the Truth benchmark, with a larger advantage when load uncertainty is amplified beyond the nominal training distribution. In both experiments, EnSF reduces load-estimation error and improves the resulting control performance in the simplified transmission network.
Related
- Deep Reinforcement Learning for Blind Passive Beamforming in RIS-Assisted MIMO Systems
- Deep Reinforcement Learning for BD-RIS-Assisted Multi-UAV RSMA Communications
- Decentralized Strategies for Finite Population LQG Social Control: A Reinforcement Learning Approach
- Boundary Control of a Modified Timoshenko-Ehrenfest Beam via Reduced-Order Reinforcement Learning
- Mean Field Reinforcement Learning: Theory and Algorithms for Mean Field Control
BibTeX
@misc{hu2026dataassimilationassisted,
title = {{Data-Assimilation-Assisted Reinforcement Learning for Power Grid Control under Load Uncertainty}},
author = {Ruoyu Hu and Feng Bao and Zezhong Zhang and Yanzhao Cao and Guannan Zhang},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.27229},
eprint = {2609.27229},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.27229},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Data-Assimilation-Assisted Reinforcement Learning for Power Grid Control under Load Uncertainty
AU - Ruoyu Hu
AU - Feng Bao
AU - Zezhong Zhang
AU - Yanzhao Cao
AU - Guannan Zhang
PY - 2026
DA - 2026/09/23
PB - arXiv
DO - 10.48550/arxiv.2609.27229
UR - https://arxiv.org/abs/2609.27229
AB - Reliable power grid control requires sequential decisions under transmission constraints, time-varying demand, renewable variability, and imperfect load information. We investigate how load information quality affects control in a customized transmission network simulator inspired by Grid2Op and based on a DC power flow model. The environment includes stochastic spatial loads, temporally correlated renewable-like generation, energy storage, line protection, generator dispatch, and line reconnection. An Ensemble Score Filter (EnSF) corrects noisy load information before it is passed to the controller. A heuristic study first isolates the effect of this correction under a fixed feedback rule. Proximal policy optimization (PPO) agents are then trained for sequential generator and reconnection control, frozen, and evaluated with Forward, EnSF-corrected, and Truth load information on paired stochastic scenarios. EnSF consistently lowers load estimation error. Under heuristic control, the correction extends survival and reduces warning and overload exposure. With frozen PPO policies, it increases cumulative return and keeps performance closer to the Truth benchmark, with a larger advantage when load uncertainty is amplified beyond the nominal training distribution. In both experiments, EnSF reduces load-estimation error and improves the resulting control performance in the simplified transmission network.
ER -