Anomaly Data of Lithium-Ion Batteries: Labeled 256-Step EV Battery Pack Windows for Anomaly Detection (Tsinghua University, Parquet conversion)
- Provider
- Dongxu Guo (Tsinghua University); Parquet conversion by Crn-acc
- License
- CC-BY-4.0 (source)
- Size
- 4 files
- Format
- Parquet (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/Crn-acc/lithium-ion-anomaly)
- Access
- Free and public. No account with the data provider is needed.
- DOI
- 10.17632/zdjymzjz6n.1
- Category
- Energy
- Source
- https://doi.org/10.17632/zdjymzjz6n.1
About this dataset
Electric vehicle (EV) battery anomaly detection data from Tsinghua University, published by Dongxu Guo on Mendeley Data as Anomaly Data of Lithium-ion Batteries. Each sample is a window of 256 time steps with eight channels from vehicle battery operation: pack voltage, current, state of charge (SOC), maximum and minimum single-cell voltage, maximum and minimum temperature, and a timestamp counter, with metadata giving vehicle mileage and, for training windows, a label marking normal or abnormal segments. The original release ships two ZIPs of Python pickle files (Train.zip, Test.zip). This Hugging Face repository converts them to two Parquet tables in long format, 256 rows per window: train.parquet with 28,389 labeled windows (23,735 normal, 4,654 anomalous; about 7.27 million rows) and test.parquet with 11,234 unlabeled windows (label -1). Columns are volt, current, soc, max_single_volt, min_single_volt, max_temp, min_temp, timestamp, file (window id), label and mileage. Load with pandas.read_parquet or the datasets library. Useful for battery management system (BMS) fault and thermal-runaway precursor detection. Original data licensed CC BY 4.0; cite Guo, Mendeley Data, doi:10.17632/zdjymzjz6n.1.
How to load this dataset in Python
from datasets import load_dataset
ds = load_dataset("Crn-acc/lithium-ion-anomaly")
print(ds)Data structure
File types in sampled files: Parquet
Files
Example file: test.parquet
Analyze this dataset without downloading it
On Scigantic, Anomaly Data of Lithium-Ion Batteries: Labeled 256-Step EV Battery Pack Windows for Anomaly Detection (Tsinghua University, Parquet conversion) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Energy datasets
- Department of Energy’s Geothermal Data Repository (GDR) Data Lake
- Smart Buildings Dataset SB1-SB11: HVAC Building Automation Time Series
- DOE Geothermal Data Repository: Seismic Survey 2016 Data at Crescent Valley, Nevada
- Severson/Attia Stanford-MIT-Toyota Battery Cycle Life Data: 124 Commercial LFP/Graphite Cells Cycled to Failure under Fast Charging (bsebench copy)
- Reduced-Form Dispatch Model of Indian Power Generation: Plant-Level CO2 and SO2 Emissions, 2017-2018 (CMU KiltHub)
- ConStellaration: Quasi-Isodynamic Stellarator Plasma Boundaries with Ideal MHD Equilibria
- DOE Geothermal Data Repository: Data Catalog of the Geothermal Data Repository (GDR)
- Kardashev Open U.S. Grid Dataset: ERCOT interconnection queues, large-load observations and fuel mix