1 College of Transportation Science and Engineering, Civil Aviation University of China, Tianjin 300300, China
2 College of Safety Science and Engineering, Civil Aviation University of China, Tianjin 300300, China
3 Key Laboratory of Civil Aviation Thermal Disaster Prevention and Emergency Response, Civil Aviation University of China, Tianjin 300300, China
Show less
文章历史+
Received
Published
Revised
2025-06-24
2026-06-08
2025-11-19
Online Date
2026-08-07
PDF (9178K)
摘要
锂离子电池热失控是亟待解决的关键安全问题,热失控早期识别与预警对防止灾难性后果至关重要。提出基于拉曼光谱气体分析的早期识别与爆炸预测算法,联立偏最小二乘回归(partial least squares regression,PLSR)与卷积神经网络(convolutional neural network,CNN),同时引入注意力机制提高对关键信息的关注度和理解力,实现对热失控特征气体(C2H4、CO、CH4等)的精准检测,通过融合热失控动力学与Raman-CNN气体体积分数预测算法,构建爆炸时刻预测模型。经验证,该模型对爆炸时间预测的平均绝对误差体现为时间提前24 s,该预警响应机制能有效降低锂电池热失控造成的人员及财产损失。
Abstract
Thermal runaway of lithium-ion batteries represents a critical safety issue that requires urgent resolution, wherein early identification and warning are considered essential for preventing severe consequences. An early identification and explosion prediction algorithm based on Raman spectroscopic gas analysis was proposed. The method integrated partial least squares regression (PLSR) with a convolutional neural network (CNN), and an attention mechanism was introduced to enhance the focus on and interpretation of key features, thereby achieving accurate detection of characteristic gases associated with thermal runaway, such as C2H4, CO, and CH4. By combining thermal runaway dynamics with the Raman-CNN gas volume fraction prediction algorithm, a model for predicting the time to explosion was constructed. Validation results indicate that the mean absolute error of the model’s predicted explosion time is 24 seconds. It is expected that this warning mechanism can effectively mitigate personnel injuries and property losses resulting from lithium battery thermal runaway.
ZHANG Wei, ZHANG Tian-long, YANG Xu, ZHANG Hai-jun.
Thermal Runaway Early Warning and Explosion Prediction in Lithium-ion Batteries with Raman-CNN[J].
Science Technology and Engineering, 2026, 26(16): 6902-6910 DOI:10.12404/j.issn.1671-1815.2504718
偏最小二乘法(partial least squares,PLS)是一种广泛应用于多元统计分析的方法,主要用于解决回归问题。通过将实验所得到的单一气体光谱数据与体积分数建立联系,构建基于偏最小二乘回归(partial least squares regression,PLSR)的定量分析模型[15];之后进行归一化处理使卷积神经网络模型数据规范化、避免拟合过度和加快收敛。将包括H2、O2、CO、CO2、CH4、C2H4、C3H6在内的不同体积分数的标准气体光谱数据作为特征光谱数据,利用PLSR搭建定量分析模型,训练集为1 500,验证集为500。经验证,该7种热失控特征气体验证集预测结果中决定系数(R2)最小为0.946,均方根误差(root mean square error,RMSE)最大仅为0.405%。表明所建立的PLSR回归模型预测精度达到满意效果,可以作为卷积神经网络分析的标签。
式(1)中:${y}_{n}^{m+1}$(τ)为第m+1层的第n个特征映射在位置τ处的输出值;σ为激活函数ReLU;${x}_{n}^{m}$为第m层中用于生成第n个特征映射的偏置项;Fm为第m层输入特征映射的总数;${K}_{nf}^{m}$(τ)为是一维卷积核,用于将第m层的第 f 个输入特征映射${y}_{f}^{m}$转换为第m+1层的第n个输出特征映射;*为一维卷积运算。
基于Raman-CNN模型预测的气体体积分数(包含CO、H2、CH4等),使用Le Chatelier混合法则计算爆炸极限,如式(5)所示。其中单独气体的爆炸极限根据美国国家消防协会(National Fire Protection Association,NFPA)发布的Standard on Explosion Protection by Deflagration Venting(NFPA 68—2022)选择。
YinYuqiang, ZhengKang, XingJiani. Case analysis and counter measure proposal of fire accidents of lithium battery energy storage station[J]. Electric Engineering, 2024(18): 70-72.
[6]
ZhangY, WangH, YuH, et al. Influence of cathode materials on the characteristics of lithium-ion battery gas generation during thermal runaway[J]. Fire Technology, 2024, 61: 3935-3954.
ZhangZhenyu, LeiDa, ZhangYifeng, et al. MoS2@c-MOF sensor for detecting thermal runaway gas C2H4 in lithium iron phosphate batteries[J]. Electronic Components and Materials, 2024, 43(9): 1096-1103.
[9]
WeiseD, FletcherH T, JohnsonJ T, et al. The composition of gases from a diffusion flame above longleaf pine needle fuel beds[J]. Fuel, 2026, 406: 136681.
CaoJiangtao, LiQuancheng, BanMing, et al. Review of hazardous gas detection technology based on infrared spectral imaging[J]. Science Technology and Engineering, 2023, 23(19): 8050-8060.
YangMengjie, YangAijun, YeYijun, et al. Research progress on early warning of thermal runaway in lithium-ion batteries based on gas analysis[J]. Transaction of China Electrotechnical Society, 2023, 38(17): 4507-4538.
YangQingying, ChengCunfeng, SunYu, et al. Cavity-enhanced Raman spectroscopy for trace hydrogen gas sensing[J]. Chinese Journal of Quantum Electronics, 2021, 38(5): 669-676.
[16]
ZhangQ, LiuT, HaoC, et al. In situ Raman investigation on gas components and explosion risk of thermal runaway emission from lithium-ion battery[J]. Journal of Energy Storage, 2022, 56:105905.
[17]
ShirmohammadM, ShortM A, ZengH S, et al. Collision enhanced Raman scattering (CERS): an ultra-high efficient raman enhancement technique for hollow core photonic crystal fiber based Raman spectroscopy gas analyzer[J]. Biosensors, 2023, 13(11): 979.
WuWenze, HeKai, WuDonglei. Quantification of adulteration in extra virgin olive oil based on Raman spectroscopyand CNN algorithm[J]. China Oils and Fats, 2024, 49(5): 133-137.
MiaoQingxuan, HuZilong, YeCheng, et al. Online detection of DHA in continuous fermentation processes based on 1D-CNN and Raman spectroscopy[J/OL]. (2025-05-13)[2025-06-24]. https://doi.org/10.16085/j.issn.1000-6613.2025-0527.
ZhangXuesong, SunMingsi, LiuHuanzheng. Quantitative analysis of avocado oil using one-dimensional convolutional neural network model and in situ micro Raman spectroscopy[J]. The Journal of Light Scattering, 2024, 36(4): 445-453.
ChenDa, HaoChaolong, LiuTiantian, et al. Research on Raman spectroscopy analysis method of thermal runaway gas in lithium ion batteries[J]. Chinese Journal of Lasers, 2022, 49(23): 186-194.
LiShiwan, YuanMingdao, XuYunqian, et al. CNN-LSTM seepage quantity prediction model of earth-rock dam based on attention mechanism[J]. Science Technology and Engineering, 2025, 25(21): 9102-9108.
[28]
ChenH, YangK, ShaoJ, et al. Explosion dynamics for thermal runaway gases of 314 Ah LiFePO4 lithium-ion batteries triggered by overheating and overcharging[J]. Process Safety and Environmental Protection, 2024, 192: 1238-1248.
[29]
WangK, WuD, ChangC, et al. Charging rate effect on overcharge-induced thermal runaway characteristics and gas venting behaviors for commercial lithium iron phosphate batteries[J]. Journal of Cleaner Production, 2024, 434: 139992.
[30]
SuL, YangF, HuW, et al. Thermal runaway gas generation of lithium iron phosphate batteries triggered by various abusive conditions[J]. Journal of Energy Engineering, 2024, 150(4): 5122.