bsebench-org/empa-aurora-2025

Empa Aurora 2025数据集是Empa / ETH Zürich / EPFL / SINTEF Aurora 2025硬币电池循环数据集的第2层规范化视图，采用BSEBench规范Parquet模式（基于BPX-1.1符号约定和BattINFO对齐的列名）。该数据集源数据已符合LF Energy电池数据联盟的电池数据格式（BDF），但通过规范化处理（包括模式重命名、符号约定调整和派生列计算）提供了统一的视图，以支持电池状态估计（如充电状态和健康状态）的基准测试。数据集包含199个电池单元的循环数据，目标架构包括cell_id、time_s、voltage_V、current_A、temperature_C、cycle_number、step_id、capacity_Ah、soc_truth和soh_truth等列，旨在促进跨数据集的集成和下游滤波器的应用。Empa Aurora 2025 dataset is the Tier 2 canonical harmonized view of the Empa / ETH Zürich / EPFL / SINTEF Aurora 2025 coin-cell cycling dataset, in the BSEBench-canonical Parquet schema (BPX-1.1 sign convention, BattINFO-aligned column names). The source dataset is already BDF-native, but the harmonization provides a unified view through schema renaming, sign convention reconciliation, and derived columns (e.g., soc_truth and soh_truth) to support battery state-estimation benchmarking. It includes data from 199 cells, with target columns such as cell_id, time_s, voltage_V, current_A, temperature_C, cycle_number, step_id, capacity_Ah, soc_truth, and soh_truth, facilitating cross-dataset integration and downstream filter applications.