BIG LEAP Project: SoC/SoE Estimator Data (NMC Baseline and NCA Transfer Learning).

# BIG LEAP - SoC/SoE Estimator Data (NMC Baseline and NCA Transfer Learning) These files contain **processed experimental cell data**, including measured operating signals, calculated SoC/SoE series, **explicit SoC/SoE reference values**, and identifiers of the individual cells. They support the joint State of Charge (SoC) and State of Energy (SoE) estimator developed by IKERLAN within **BIG LEAP** (Horizon Europe Grant Agreement **101137815**). The estimator and its assessment are documented in **D3.2, _Development of data-driven SoX and RUL algorithms for increase BMS interoperability_ (30 June 2025)**: Sections **1.1.2-1.1.3** describe the model design and NMC baseline, while Sections **1.1.4 and 1.1.4.1** describe the transfer-learning approach and NCA evaluation. ## Included files | File | Chemistry | Purpose | Data rows | Distinct cells || --- | --- | --- | ---: | ---: || `Baseline_NMC_testing_cells_data.csv` | Nickel manganese cobalt (NMC) | Processed data associated with the **baseline** SoC/SoE estimator | 180,771 | 2 || `TL_NCA_testing_cells_data.csv` | Nickel cobalt aluminium (NCA) | Processed data associated with **transfer learning** from NMC to NCA | 703,977 | 7 | ## Column definitions | Column | Description | Units / format || --- | --- | --- || `Chemistry` | Battery chemistry (`NMC` or `NCA`). | Text || `Cell_ID` | Identifier of the physical test cell associated with the sample. Use to group samples by cell and distinguish different tests/replicates. | Text || `time` | Time increment between recorded samples (delta-t), **not** a global timestamp. | Seconds || `Voltage` | Measured cell voltage. | V || `Current` | Measured cell current; the files use positive current for charging and negative current for discharging. | A || `Temperature` | Measured cell temperature. | Degrees Celsius || `SoH` | Cell State of Health associated with the sample. | Fraction of reference capacity (nominally 0-1) || `SoC_calc` | Calculated SoC series included in the processed export. It is **distinct from** `SoC_reference`; its exact computational origin is not identified by the CSV header alone. | Fraction (1 = 100%) || `SoE_calc` | Calculated SoE series included in the processed export. It is **distinct from** `SoE_reference`; its exact computational origin is not identified by the CSV header alone. | Fraction (1 = 100%) || `SoC_reference` | Explicit reference SoC value for the sample, to be distinguished from the calculated SoC series. | Fraction (1 = 100%) || `SoE_reference` | Explicit reference SoE value for the sample, to be distinguished from the calculated SoE series. | Fraction (1 = 100%) | **Cell identification:** Each row carries its own `Cell_ID`. For example, `SNL_18650_NMC_25C_0-100_0.5-2C_a` identifies one of the NMC cells; the coded name contains nominal test conditions and a replicate suffix. This enables cell-level filtering and visualization without confusing observations from different cells. There are **2 distinct NMC IDs** and **7 distinct NCA IDs** in the supplied CSVs. **Reference values:** Reference states for SoC/SoE training and evaluation were derived from experimental data using coulomb counting with periodic recalibration after full-charge cycles. The `SoC_reference` and `SoE_reference` columns explicitly distinguish the reference quantities from the `*_calc` quantities. ## Practical notes for reuse - **Group by `Cell_ID`** before plotting cell-specific state trajectories or calculating statistics; neither CSV represents one uninterrupted battery history across all rows.- Treat `time` as the interval between samples. Do not read it as an absolute time axis or combine elapsed times across different cells. - Some values of `SoH`, `SoC_calc`, `SoE_calc`, `SoC_reference`, and `SoE_reference` lie slightly outside their nominal **0-1** range. Preserve the raw values unless a documented cleaning rule calls for clipping/filtering. ## Source dataset references **Sandia National Laboratories - experimental NMC and NCA cell cycling data:** Y. Preger et al., "Degradation of Commercial Lithium-Ion Cells as a Function of Chemistry and Cycling Conditions," *Journal of The Electrochemical Society*, **167**(12), 120532 (2020). DOI: [10.1149/1945-7111/abae37](https://doi.org/10.1149/1945-7111/abae37). - [Battery Archive - Sandia study and data](https://www.batteryarchive.org/snl_study.html)- [Sandia National Laboratories publication record](https://www.sandia.gov/research/publications/details/degradation-of-commercial-lithium-ion-cells-as-a-function-of-chemistry-and-2020-01-09/)- The study is cited as **reference [2]** in BIG LEAP deliverable D3.2. **BIG LEAP project report:** *D3.2 - Development of data-driven SoX and RUL algorithms for increase BMS interoperability (1st period report)*, 30 June 2025, Section **1.1**, particularly **1.1.2**, **1.1.3**, and **1.1.4.1**.