empa-aurora-2025
收藏资源简介:
Empa Aurora 2025数据集是Empa、ETH Zürich、EPFL和SINTEF Aurora 2025纽扣电池循环数据集的Tier 2规范化视图,采用BSEBench-canonical Parquet模式(BPX-1.1符号约定,BattINFO对齐的列名)。该数据集包含199个电池单元的时间序列数据,原始数据已采用LF Energy Battery Data Alliance的电池数据格式(BDF)。数据集的目标是提供跨数据集统一的规范化视图,便于电池状态估计滤波器的基准测试。数据集包含10个核心列:cell_id(电池标识符)、time_s(时间,秒)、voltage_V(电压,伏特)、current_A(电流,安培,采用BPX-1.1符号约定:充电为正,放电为负)、temperature_C(温度,摄氏度)、cycle_number(循环编号)、step_id(步骤标识符)、capacity_Ah(容量,安时)、soc_truth(真实荷电状态,0-1)、soh_truth(真实健康状态,0-1)。数据集还包含每个电池的元数据侧车文件,包括化学系列、标称容量等。数据集适用于时间序列预测和表格回归任务,特别是电池状态估计(如SOC和SOH)的基准测试。数据集采用Creative Commons Attribution 4.0 International (CC-BY-4.0)许可。
The Empa Aurora 2025 dataset is a Tier 2 canonical view of the Empa, ETH Zürich, EPFL, and SINTEF Aurora 2025 coin cell cycling dataset, using the BSEBench-canonical Parquet schema (BPX-1.1 sign convention, BattINFO-aligned column names). It contains time-series data from 199 battery cells, with raw data formatted according to the LF Energy Battery Data Alliances Battery Data Format (BDF). The dataset aims to provide a unified canonical view across datasets for benchmarking battery state estimation filters. It includes 10 core columns: cell_id (cell identifier), time_s (time in seconds), voltage_V (voltage in volts), current_A (current in amperes, following BPX-1.1 sign convention: positive for charging, negative for discharging), temperature_C (temperature in degrees Celsius), cycle_number (cycle number), step_id (step identifier), capacity_Ah (capacity in ampere-hours), soc_truth (true state of charge, 0-1), soh_truth (true state of health, 0-1). The dataset also includes metadata sidecar files for each cell, covering chemical series, nominal capacity, etc. It is suitable for time-series forecasting and tabular regression tasks, particularly for benchmarking battery state estimation (e.g., SOC and SOH). The dataset is licensed under Creative Commons Attribution 4.0 International (CC-BY-4.0).




