akiroussama's picture
P1.6: add Tier-1 README documenting 198 .mat + 59 EIS .csv + 2 metadata (K-021 closure)
0f3c966 verified

Panasonic NCR18650PF — Tier-1 raw mirror

BSEBench Tier-1 mirror of the Panasonic NCR18650PF battery testing data published by Phillip Kollmeyer (2018) on Mendeley Data.

Original source

  • Publication : Kollmeyer, P. (2018). "Panasonic 18650PF Li-ion Battery Data and Experimental Procedures." Mendeley Data, v1.
  • DOI : 10.17632/wykht8y7tg.1
  • Upstream URL : https://data.mendeley.com/datasets/wykht8y7tg/1
  • License : CC-BY-4.0
  • Cell : Panasonic NCR18650PF (NCA chemistry, 18650 cylindrical, rated 2.9 Ah)

File breakdown (260 total)

Type Count Purpose
.mat 198 Time-series measurements (current, voltage, temperature, time) for each (profile, temperature) tuple, plus subsegments
.csv (EIS/) 59 Electrochemical impedance spectroscopy sweeps at multiple SOC points and temperatures
Plot_results.m 1 MATLAB utility for plotting (upstream-provided)
Readme file - desc of tests performed.txt 1 Upstream documentation of the experimental procedure
README.md 1 This file (BSEBench-added)

Folder layout (preserves upstream hierarchy)

Panasonic 18650PF Data/
├── -20degC/  ├── EIS/        # impedance sweeps
│              └── *.mat       # drive-cycle measurements
├── -10degC/  ...
├── 0degC/    ...
├── 10degC/   ...
├── 25degC/   ...
├── 10degC Trise with pause/   # thermal characterization
├── -20degC Trise/             # thermal characterization
└── Plot_results.m

Profiles measured

DST, US06, FUDS, HWFET, LA92, NN (Neural Network — Kollmeyer custom), Cycle_1..Cycle_4 — at 5 temperatures (-20, -10, 0, 10, 25 °C).

BSEBench Tier-2 (harmonized)

The corresponding harmonized canonical Parquet files (BSEBench v1.1.0 schema) are at bsebench-org/panasonic-kollmeyer-2018 (Tier-2 repo). See bsebench-runner/docs/M1_HF_LAYOUT.md for the convention.

K-016 closure note

The BSEBench M1 adapter adapters/panasonic_kollmeyer_2018.py produces Tier-2 Parquet that pad cell_id and cycle_number as constant columns (single-cell driving-cycle dataset). See KAIZEN K-016 for the rationale.