基于Transformer神经网络的锂电池热失控多数据融合探测
Thermal runaway multi-data fusion detection of lithium battery based on Transformer neural network
丁沐涛 1郭世伟 1单志林 1张启兴1
作者信息
- 1. 中国科学技术大学 火灾科学国家重点实验室,安徽 合肥 230026
- 折叠
摘要
为满足对锂离子电池热失控高效准确探测的需求,设计了一种锂电池热失控试验平台,并利用STM32F103ZET6单片机连接了CO、CO2、H2 和热敏电阻NTC共 4种传感器,实时采集特征参量.同时,利用PyroSim模拟试验环境,生成高质量的模拟数据,以补充试验数据.基于PyTorch平台,设计了一个Transformer神经网络,能够输出锂电池的正常、预警和热失控 3种状态.通过使用试验数据和模拟数据进行训练,实现了对锂电池热失控的融合探测,相比于其他算法有一定的优势.
Abstract
In order to meet the demand for efficient and accurate detection of lithium ion battery thermal runaway,this study de-signed a lithium battery thermal runaway experimental platform.STM32F103ZET6 single chip microcomputer was used to con-nect four sensors such as carbon monoxide,carbon dioxide,hy-drogen and NTC to collect characteristic parameters in real time.At the same time,PyroSim is used to simulate the experimental environment and generate high-quality simulation data to supple-ment the experimental data.Based on the pytorch platform,we designed a Transformer neural network that can output the nor-mal,early warning and thermal runaway states of lithium batter-ies.By using experimental data and simulation data for training,we successfully achieved fusion detection of thermal runaway data of lithium batteries,and verified the effectiveness of the algorithm.
关键词
热失控/特征参量/PyroSim/PyTorch/Trans-former/数据融合Key words
thermal runaway/characteristic parameter/PyroSim/PyTorch/Transformer/data fusion引用本文复制引用
基金项目
国家重点研发计划课题(2021YFC3001601)
出版年
2024
NETL
NSTL
万方数据