Multi-Objective Graph-Based Reinforcement Learning for Voltage and Frequency Stability in Grid-Connected Renewable Energy Systems
Bibliographic record
Abstract
The increasing penetration of renewable energy sources in modern power grids poses significant challenges in maintaining voltage stability, frequency regulation, and optimal power dispatch due to the stochastic nature of generation and demand. This paper introduces a multi-objective graph-based reinforcement learning (MO-GRL) framework that optimally manages energy storage systems (ESS) to enhance voltage stability, provide synthetic inertia for frequency support, and maximize energy efficiency. The proposed model integrates Koopman operator theory for predictive ESS dynamics, graph neural networks for learning spatial relationships in power networks, and soft actor-critic reinforcement learning for multi-objective optimization. Furthermore, a Lyapunov-based stability constraint is incorporated to ensure the robustness of voltage and frequency control policies under uncertain conditions. Compared to conventional techniques, MO-GRL dynamically adjusts control policies based on real-time grid states, offering improved adaptability. Simulations on an IEEE 33-bus test system demonstrate significant improvements, including a 42% reduction in voltage deviations, a 39% decrease in frequency deviations, and a 37% improvement in ESS utilization efficiency. The proposed method effectively mitigated voltage spikes that previously rose above 1.04 p.u., successfully limiting peak voltage levels to 1.03 p.u. and lower. Similarly, during peak evening demand conditions, the voltage dips that fell below 0.95 p.u. were stabilized above 0.96 p.u. The controlled voltage profile-maintained values near the desired reference of 1.00 p.u., ensuring improved voltage stability across the 24-hour operation.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".