Privacy-Preserving Federated Reinforcement Learning for BESS Coordination in Distribution Networks with Voltage Regulation
Bibliographic record
Abstract
The increasing integration of distributed energy resources (DERs) in distribution networks, particularly battery energy storage systems (BESSs), enables energy trading and operational cost reduction but also introduces challenges for coordinated energy management. Effective coordination of distributed BESSs requires access to residential data, raising significant data privacy concerns. In recent years, federated reinforcement learning (FRL) has emerged as a promising AI approach for privacy-preserving BESS energy management. However, the application of FRL in distribution networks is limited by several challenges, including coordination under voltage constraints caused by coupled BESS interactions, privacy leakage from intermediate result sharing, and scalability issues under high BESS penetration.To address these challenges, this paper proposes a privacy-preserving federated reinforcement learning scheme for BESS coordination with voltage regulation (PPFRL-BC). The BESS energy management problem is formulated under (ε,δ)-differential privacy constraints, within which privacy-preserving reward functions jointly capture system-level rewards and voltage regulation requirements. Building on this formulation, the proposed PPFRL-BC scheme enables coordinated BESS operation without direct sharing of raw residential data, while a BESS-oriented Gaussian noise mechanism is incorporated to mitigate privacy leakage associated with intermediate result sharing. To further improve scalability, an approximate action branching method is introduced to reduce computational complexity in large-scale distribution networks.Case studies on the IEEE 33-bus and 123-bus test feeders demonstrate that the proposed PPFRL-BC scheme effectively coordinates distributed BESS operations, maintains voltage regulation under network constraints and preserves data privacy, while achieving performance comparable to the centralized scheme and outperforming both decentralized and conventional horizontal FRL schemes.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.010 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".