Dataset for : Realistic Multi-Fault Diagnostics of Millions-Scale Li-ion batteries with Rapid Unsupervised Learning 核心 · 已核验

doi:10.5281/zenodo.18328700

Dataset for : Realistic Multi-Fault Diagnostics of Millions-Scale Li-ion batteries with Rapid Unsupervised Learning

Abstract

The rapid deployment of battery swapping stations necessitates scalable and reliable fault diagnosis, yet massive, sparse operational data and scarce labeled samples make this challenging. Here, we report a rapid unsupervised learning framework for realistic  multi-fault diagnosis in million-scale battery fleets. Our approach employs a double-layer mechanism. First, we rapidly screen for abnormal devices by extracting features from voltage-envelope sequences. Subsequently, we pinpoint faulty cells and types using an enhanced two-stage unsupervised clustering combined with rule-based fault tracing. The framework is validated on a production dataset of over 128,000 devices, achieving 97.33% device-layer and 99.66% cell-layer accuracy. Laboratory tests on recalled batteries further confirm the detection of low-capacity and micro-short-circuit faults. These results demonstrate scalability and robustness under sparse-data conditions, enabling reliable operations for large-scale energy storage systems.

Dataset Structure

DataRepo/

├── fullDataset/

│   └── fullDataset.json             # Feature data for all devices

└── predefinedDataset/

├── data/                        # Raw data for predefined devices

└── processedData/

├── predefinedFeatures.json  # Extracted features for predefined devices

├── device_level_info.csv    # Device-level information extracted from JSON

└── cell_level_info.csv      # Cell-level information extracted from JSON

Dataset Description

1. Predefined Dataset (predefinedDataset/)

The predefined dataset contains data for a selected set of devices used in preliminary research:

Raw Data ( data/ )

Contains raw voltage data files for predefined devices

Each file represents voltage measurements from a single device

Data format: CSV files with timestamp and voltage readings fo

落地页

https://zenodo.org/doi/10.5281/zenodo.18328700

许可证

CC-BY-4.0 (判读置信:inferred)

国内可访问性

国内直连:可达 (2026-07-11 检测) 代理通道:可达 (2026-07-11 检测)
检测口径:lychee 双通道单轮探测;「直连超时」表示检测窗口内未完成,系慢或不稳定证据,不构成封锁证据。

发布年份

2026

发布方

Zenodo

设备类型

battery

PHM 任务

fault_diagnosis fault_detection

故障工况