Centralized Multi-Agent SOC Control for Battery Health Using Proximal Policy Optimization in EVs
Bibliographic record
Abstract
Lithium-ion batteries (LIBs) have garnered significant attention due to their expanding use in various applications, including electric vehicles (EVs) and smart grids. To meet the diverse requirements of these applications, LIB cells are configured in different architectures, such as multiple cell/module/pack which are arranged in series and parallel configurations. In series configurations, a state of charge (SOC) balancing system is essential to ensure uniform SOC levels across all cells. For battery electric vehicles (BEVs), which rely solely on LIBs as their energy storage system (ESS), maximizing the ESS capacity is crucial for extending the driving range. SOC balancing is a key strategy to achieve optimal utilization of ESS capacity in EVs. This paper presents a model-free cooperative multi-agent control framework designed to regulate and balance the SOC of lithium-ion battery (LIB) cells in EVs during real-time driving operations. The proposed method utilizes a series architecture comprising three LIB cells, each equipped with a buck-boost converter and a proportional-integral (PI) controller, controlled by a reinforcement learning (RL) agent. The Proximal Policy Optimization (PPO) algorithm is used as the RL agent in this multi-agent framework, where each PPO agent independently manages the SOC of a corresponding battery cell based on observed data. During the training phase, all PPO agents work collaboratively to balance the SOCs of the LIB cells, thereby preventing interruptions in EV performance. The effectiveness of the proposed approach is demonstrated by comparing its performance with single-agent methods such as PPO, Soft Actor-Critic (SAC), and Twin Delayed Deep Deterministic Policy Gradient (TD3), as well as with other multi-agent methods. The results show that the proposed method performs better than the existing approaches, indicating its potential for superior performance.
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 | 0.001 |
| 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 | 0.001 | 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".