Python implementation of battery Equivalent Circuit Models (Thevenin / Extended Thevenin) for SOC, SOH, and RUL estimation. Includes cell-level and pack-level simulation with passive/active cell balancing. Based on academic research at TU Berlin and coursework from University of Colorado Boulder.
This project provides a complete battery simulation framework that:
- Models individual cells using 1-RC and 2-RC Thevenin ECMs
- Estimates State of Charge (SOC) via Coulomb counting + OCV correction
- Estimates State of Health (SOH) from capacity fade and resistance growth
- Predicts Remaining Useful Life (RUL) using degradation models
- Simulates full battery packs (14S3P) with cell-level parameter variation
- Implements passive and active cell balancing strategies
- Includes thermal modeling (temperature-dependent resistance)
battery-ecm-simulation/
├── src/
│ ├── ecm_model.py # Core ECM: cell model, SOC/SOH/RUL, current profiles
│ ├── pack_simulation.py # Multi-cell pack + cell balancing
│ └── visualize.py # Plotting + simulation reports
├── notebooks/
│ └── ecm_analysis.ipynb
├── data/
├── results/
├── requirements.txt
└── README.md
git clone https://github.com/PRATdoppelEK/battery-ecm-simulation.git
cd battery-ecm-simulation
pip install -r requirements.txtfrom src.ecm_model import BatteryECM, CellParameters, cc_discharge_profile
params = CellParameters(capacity_ah=50.0, r0=0.002, r1=0.001, c1=3000)
cell = BatteryECM(params=params, extended=True, dt=1.0)
cell.reset(initial_soc=1.0)
profile = cc_discharge_profile(50, c_rate=0.5)
history = cell.simulate(profile)from src.pack_simulation import BatteryPackSimulator, PackConfiguration
config = PackConfiguration(n_series=14, n_parallel=3, capacity_ah=50)
pack = BatteryPackSimulator(config)
pack.reset(initial_soc=0.95, soc_spread=0.05)
pack.simulate_pack(profile, balance=True)python src/visualize.py
| Feature | Description |
|---|---|
| ECM Type | 1-RC and 2-RC Thevenin |
| SOC method | Coulomb counting + OCV lookup |
| Thermal model | Lumped thermal mass, ambient coupling |
| Balancing | Passive (resistive bleed) + Active (charge transfer targets) |
| Current profiles | CC-Discharge, CCCV-Charge, WLTP-like drive cycle |
| SOH model | Capacity fade + resistance growth |
| RUL prediction | Linear degradation extrapolation |
- OCV–SOC lookup table for NMC chemistry (interpolated)
- Discrete-time RC dynamics: exact exponential integration (no Euler approximation errors)
- Temperature-dependent R0: linear correction around 25°C reference
- Cell SOH variation: random spread across pack cells for realistic imbalance simulation
- Validated against MATLAB/Simulink reference models (TU Berlin thesis, Dan-Tech Energy)
- M.Sc. Thesis: "Custom Battery Cell Balancing Circuit Design Under Thermal Gradient" — TU Berlin
- Certifications: Battery Management Systems (with honors), Equivalent Circuit Cell Model Simulation, Battery Pack Balancing and Power Estimation — University of Colorado Boulder
NumPy · Matplotlib · SciPy · Pandas · Jupyter · Python 3.10+
Prateek Gaur — ML Engineer | Battery & Engineering AI
LinkedIn · GitHub