3S LiPo Battery Pack Thevenin ECM (12-state)
ADVANTAGES1 · dim 12SolvSRK wins. At the comparison noise level, SolvSRK beats the best baseline by at least 10 percentage points of survival, or by at least 0.05 balanced score when survival is tied. Use SolvSRK for this class of problem. All verdicts →
12-state 3S LiPo battery pack discharge: 2nd-order Thevenin equivalent circuit model (Chen & Rincon-Mora 2006). 4 states per cell (SOC, V_rc1, V_rc2, T_cell) × 3 cells in series. Piecewise-constant mission current profile. tau_RC1 (~10s) vs tau_thermal (~300s) → stiffness ~30:1.
Problem definition
Chen & Rincon-Mora, 'Accurate Electrical Battery Model Capable of Predicting Runtime and I-V Performance', IEEE Trans. Energy Conversion, vol. 21, 2006
Canonical RHS excerpt from the registered callable used for this benchmark cell. Expand it to verify the state equations; it is not a standalone runnable fixture.
Show canonical RHS excerpt
def _mission_current(t: float) -> float:
"""Pack current draw (A) for a typical survey mission.
0–60 s: hover (15 A)
60–300 s: cruise (8 A)
300–360 s: hover (15 A)
360–420 s: descent (5 A)
420+ s: idle (0.5 A)
"""
if t < 60.0:
return 15.0
elif t < 300.0:
return 8.0
elif t < 360.0:
return 15.0
elif t < 420.0:
return 5.0
else:
return 0.5
def _r0_effective(soc: float, temp: float, variability: float) -> float:
"""R0 as a function of SOC, temperature, and cell variability."""
soc_factor = 1.0 + _R0_SOC_FACTOR * (1.0 - np.clip(soc, 0.0, 1.0))
temp_factor = 1.0 + _ALPHA_R * (temp - 25.0)
return _R0_BASE * soc_factor * temp_factor * variability
def rhs_battery_3s(t, y):
"""3S LiPo pack: 12-state 2nd-order Thevenin ECM with thermal coupling."""
dy = np.zeros(12)
i_pack = _mission_current(t)
for cell in range(3):
base = cell * 4
soc = y[base]
v_rc1 = y[base + 1]
v_rc2 = y[base + 2]
t_cell = y[base + 3]
r0 = _r0_effective(soc, t_cell, _R0_VARIABILITY[cell])
# SOC dynamics: dSOC/dt = -I / Q
dy[base] = -i_pack / _Q_COULOMBS
# RC pair dynamics
dy[base + 1] = (i_pack / _C1) - (v_rc1 / _TAU1)
dy[base + 2] = (i_pack / _C2) - (v_rc2 / _TAU2)
# Thermal: lumped node with ohmic + polarisation heat generation
p_ohmic = i_pack**2 * r0
p_rc1 = v_rc1**2 / _R1
p_rc2 = v_rc2**2 / _R2
p_total = p_ohmic + p_rc1 + p_rc2
q_conv = (t_cell - _T_AMB) / _R_TH_CONV
dy[base + 3] = (p_total - q_conv) / _C_TH
return dy- Parameters
- _ALPHA_R = 0.004
- _C1 = 1000
- _C2 = 40000
- _C_TH = 45
- _Q_COULOMBS = 9000
- _R0_BASE = 0.025
- _R0_SOC_FACTOR = 0.3
- _R0_VARIABILITY = [0.95, 1, 1.05]
- _R1 = 0.01
- _R2 = 0.005
- _R_TH_CONV = 12
- _TAU1 = 10
- _TAU2 = 200
- _T_AMB = 25
- Initial condition
- y(0) = [1, 0, 0, 25, 1, 0, 0, 25, 1, 0, 0, 25]
- Horizon
- t ∈ [0, 600]
Canonical RHS excerpt captured from the same registered callable used for the published benchmark. Frozen closure values are summarized below; helper imports and solver settings are intentionally omitted.
Fingerprint
Spread: low
Default noise: low
Recommendation snapshot
Clean best: SolvSRK
Noisy best: SolvSRK
Coverage
14 solver arms · clean + 5 noise levels
Ranked on survival, precision, and speed
Versions & freeze
Methodology →- Freeze
- 2026-08-13
- libsolvsrk
- 2.3.0
- SciPy
- 1.14
- SUNDIALS
- CVODE (bundled backend)
20 seeds/cell default · 14 arms · TRL 4–5 · simulation-lab validated · this page: 3S LiPo Battery Pack Thevenin ECM (12-state) (3s-lipo-battery-pack-thevenin-ecm-12-state)
Governed SolvTune benchmark freeze; per-arm medians only. RHS definitions and raw trial rows are not published.
Self-reported by Resonix Labs · not independently verified
Results matrix
Pick an objective and a noise level to rank all arms on survival, median SCD, median nfev, and median wall time. Medians across seeds.
Objective
Best overall trade-off of survival, precision, and speed.
Noise level
| # | Solver | Survival | SCD | nfev | Wall | Score |
|---|---|---|---|---|---|---|
| 1 | SolvSRK | 100% | 10.6 | 2,796 | 52 ms | 0.872 |
| 2 | SciPy LSODASciPy | 100% | 8.6 | 877 | 14 ms | 0.823 |
| 3 | SciPy RadauSciPy | 100% | 8.2 | 3,348 | 111 ms | 0.815 |
| 4 | FBDFexternal | 100% | 8.1 | 1,024 | 5.85 s | 0.813 |
| 5 | SciPy RK45SciPy | 100% | 7.8 | 1,490 | 28 ms | 0.805 |
| 6 | SciPy RK23SciPy | 100% | 7.8 | 2,390 | 49 ms | 0.804 |
| 7 | Vern9external | 100% | 7.7 | 5,474 | 4.65 s | 0.803 |
| 8 | SciPy DOP853SciPy | 100% | 7.6 | 2,246 | 40 ms | 0.801 |
| 9 | Vern7external | 100% | 7.5 | 3,252 | 4.07 s | 0.799 |
| 10 | SciPy BDFSciPy | 100% | 7.5 | 1,462 | 65 ms | 0.797 |
| 11 | Tsit5external | 100% | 7.3 | 1,800 | 910 ms | 0.794 |
| 12 | CVODE BDFexternal | 100% | 7.0 | 993 | 28 ms | 0.785 |
| 13 | CVODE Adamsexternal | 100% | 5.5 | 718 | 22 ms | 0.751 |
| 14 | TRBDF2external | 100% | 5.1 | 3,922 | 5.82 s | 0.740 |
At Clean, best balanced arm is SolvSRK.
Values are medians across seeds, measured by Resonix Labs on Resonix hardware and not independently verified; nfev and wall are on reference lab hardware (indicative). Under injected noise only SolvSRK and the SciPy arms are run. How we measure accuracy → · Verification status →
Cite this page
Replace the access date. Pin the freeze ID and library versions when comparing against a later export. Cite it as what it is - a self-reported vendor benchmark, not an independently verified result. The note field says so; please keep it.
@misc{resonix_evidence_3s_lipo_battery_pack_thevenin_ecm_12_state_2026,
title = {Resonix Evidence Portal: 3S LiPo Battery Pack Thevenin ECM (12-state)},
author = {{Resonix Labs (Canada) Inc.}},
year = {2026},
howpublished = {\url{https://resonixusa.com/evidence/problems/3s-lipo-battery-pack-thevenin-ecm-12-state}},
note = {Self-reported vendor benchmark; internally generated by Resonix Labs and not independently verified. Accessed YYYY-MM-DD. Freeze 2026-08-13; libsolvsrk 2.3.0; SciPy 1.14.}
}Related
TRL 4–5 · simulation-lab validated · 398 problems · 14 solver arms · clean + 5 noise levels
Freeze: 2026-08-13 · scipy 1.14 · libsolvsrk 2.3.0 · Methodology
Self-reported by Resonix Labs · not independently verified · Verification status