This repository provides an advanced simulation environment for active battery cell balancing using a reinforcement learning framework. It implements a custom Gymnasium environment that models the dynamics of a lithium-ion battery pack, focusing on energy transfer between adjacent cells to achieve optimal voltage and state-of-charge (SOC) balance.
- Realistic Battery Modeling:
Simulates individual cell properties including capacity, internal resistance, and voltage-SOC mapping based on experimental Li-NMC data. - Driving Cycle Integration:
Converts real-world driving cycle data into discharge current profiles, integrating vehicle dynamics and load conditions. - Discrete Action Space:
Offers a multi-dimensional discrete action space where each action controls energy transfer intensity and direction between adjacent cells. - Custom Reward Function:
Incentivizes effective cell balancing by rewarding the minimization of SOC imbalances. - MDP Framework:
The environment is modeled as a Markov Decision Process (MDP) with well-defined state and action spaces, transition dynamics, and long-term discounting.
- Clone the Repository:
git clone https://github.com/messlem99/Battery_Cell_Balancing.git cd Battery_Cell_Balancing
To test the simulation environment, follow these steps:
- Initialize the Environment:
import gymnasium as gym from Battery_Cell_Balancing import BatteryBalancingEnv env = BatteryBalancingEnv(driving_cycle_filepath="data.csv") observation, info = env.reset(seed=42) - Run a Simulation Episode:
done = False while not done: action = env.action_space.sample() # Sample random actions observation, reward, done, truncated, info = env.step(action) print(f"Reward: {reward}, Done: {done}")
The environment is modeled as an MDP with the following components:
- State Space: Each state includes per-cell voltage, SOC values, and the instantaneous load current derived from the driving cycle.
- Action Space: A discrete multi-dimensional vector controlling energy transfers between adjacent cells.
- Transition Dynamics: Updates based on realistic battery dynamics including load current effects and energy balancing operations.
- Reward Function: Rewards effective balancing by minimizing SOC differences between adjacent cells.
- Termination Conditions: Episodes end when voltage or SOC imbalances exceed defined thresholds, or if all cells are fully discharged.
Contributions and enhancements are welcome. To get started:
- Fork the repository and create your feature branch.
- Submit pull requests for review.
- This project is licensed under the MIT License. See the LICENSE file for details.
To cite this project in publications:
@misc{EVBatteryBalancing2025,
author = {Abdelkader Messlem and Youcef Messlem and Ahmed Safa},
title = {A Simulation Environment for Active Battery Cell Balancing Using Gymnasium},
year = {2025},
howpublished = {\url{https://github.com/messlem99/Battery_Cell_Balancing}},
}- F. Baronti, W. Zamboni, N. Femia, R. Roncella, and R. Saletti, “Experimental analysis of open-circuit voltage hysteresis in lithiumiron-phosphate batteries,” in IECON 2013 - 39th Annual Conference of the IEEE Industrial Electronics Society. Vienna, Austria: IEEE, Nov. 2013, pp. 6728–6733. [Online]. Available: http://ieeexplore.ieee. org/document/6700246/
- P. Ramadass, B. Haran, R. White, and B. N. Popov, “Mathematical modeling of the capacity fade of Li-ion cells,” Journal of Power Sources, vol. 123, no. 2, pp. 230–240, Sep. 2003. [Online]. Available: https://linkinghub.elsevier.com/retrieve/pii/S0378775303005317
- G. Ning, B. Haran, and B. N. Popov, “Capacity fade study of lithium-ion batteries cycled at high discharge rates,” Journal of Power Sources, vol. 117, no. 1-2, pp. 160–169, May 2003. [Online]. Available: https://linkinghub.elsevier.com/retrieve/pii/S0378775303000296