A Model-Free Multi-Agent Reinforcement Learning Approach to Reach a Robust, Optimal, and Environment-Friendly Power Management in a Micro-Grid
Bibliographic record
Abstract
This study recommends employing recurrent neural networks and reinforcement learning-based approaches to address challenges in micro-grid (MG) power management. The proposed MG contains combined cooling, heating, and power (CCHP), wind turbine (WT), photovoltaic (PV), and battery energy storage system (BESS). To effectively estimate the production of 24-hour renewable resources, a multi-layer recurrent neural network (MLRNN) model is developed, trained by historical wind and solar variables dataset. Over previously researched methods, the approach offers advantages, including improved accuracy and extracting non-linear mapping function amongst the features and output. Moreover, the model's hyperparameters are tuned through an auto grid-search algorithm, which leads to a more accurate prediction. The forecasted values, then is employed in next study phase which is optimal power management within MG's generation units. In this regard, a model free reinforcement learning is adopted to keep the multi-objective fuel and CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> emission cost function minimized, to have a robust and environmentally friendly MG. So, a multi-agent reinforcement learning (MARL) is proposed to take the responsibility. The main objective of multi-agent solution is to overcome the challenging issue of finding the optimum operation points in a complicated environment with only one agent, by combining the simpler tasks that each agent does. According to the findings, the proposed MARL successfully dispatch the power generation among renewable resources and micro turbine in CCHP system, optimally. Moreover, the results validate the BESS proper performance in light and heavy load periods in the MG.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".