基于反熵权法的钠离子电池过充热失控多阶段预警方法研究

doi:10.19799/j.cnki.2095-4239.2025.0592

摘要/Abstract

摘要：

钠离子电池因其资源丰富、成本低和环境友好等优势，成为新型储能领域的重要发展方向。然而在高压过充等极端工况下，钠离子电池易发生热失控，严重威胁系统运行安全。本文以185 Ah层状氧化物正极钠离子单体电池为研究对象，设计过充热失控实验平台，采集电压、温度和膨胀力三类关键参数，分析其在热失控过程中的演化趋势与特征响应。基于多参量信号变化率构建动态风险指标体系，引入反熵权法实现特征权重的自适应调整，构建可随时间演化的综合风险指标。结合热失控机理划分过程阶段，提出一套多级预警模型。实验结果表明，该方法能有效识别电池在不同阶段的热失控风险，具备良好的提前预警能力与工程应用价值，可为钠离子电池的热安全管理与智能监测提供理论依据与实践支撑。

关键词: 钠离子电池, 热失控, 反熵权法, 多参量融合, 多阶段预警

Abstract:

Sodium-ion batteries have emerged as a key direction for new energy storage because of their abundant resources, low cost, and environmental friendliness. However, under extreme conditions, such as high-voltage overcharging, sodium-ion batteries are prone to thermal runaway, posing a severe threat to system operation safety. This study focuses on a 185 Ah layered oxide-positive electrode sodium-ion single-cell battery. An experimental platform was designed to simulate the overcharge-induced thermal runaway, collecting three key parameters: voltage, temperature, and pressure. The evolution trend and characteristic response of these parameters during the thermal runaway process were analyzed. A dynamic risk index system was then constructed based on the rate of change of multiple signals. To adaptively adjust feature weights, the anti-entropy weight method was introduced, leading to the development of a comprehensive risk index that evolves over time. A multi-level warning model is proposed by dividing the process stages based on the mechanism of thermal runaway. The experimental results show that the proposed method can effectively identify the thermal runaway risk of batteries at different stages and provides a strong early-warning capability with engineering application value. These findings offer a theoretical basis and practical support for the thermal safety management and intelligent monitoring of sodium-ion batteries.

Key words: sodium-ion battery, thermal runaway, anti-entropy weight method, multi-parameter fusion, multi-stage warning

中图分类号:

TM 911

吴静云, 郭鹏宇, 黄峥. 基于反熵权法的钠离子电池过充热失控多阶段预警方法研究[J]. 储能科学与技术, 2025, 14(11): 4360-4369.

Jingyun WU, Pengyu GUO, Zheng HUANG. Research on multi-stage warning method for overcharging thermal runaway of sodium-ion batteries based on anti-entropy weight method[J]. Energy Storage Science and Technology, 2025, 14(11): 4360-4369.

图/表 11

表1

图1

表2

图2

图3

图4

表3

图5

图6

图7

表4

参考文献 18

[1]	李慧, 吴川, 吴锋, 等. 钠离子电池: 储能电池的一种新选择[J]. 化学学报, 2014, 72(1): 21-29.
	LI H, WU C, WU F, et al. Sodium ion battery: A promising energy-storage candidate for supporting renewable electricity[J]. Acta Chimica Sinica, 2014, 72(1): 21-29.
[2]	应建国. 钠锂离子电池联合储能系统研究[J]. 电力电子技术, 2025, 59(4): 58-61. DOI: 10.20222/j.cnki.cn61-1124/tm.2025.04.023.
	YING J G. Research on sodium-ion and lithium-ion batteries hybrid energy storage system[J]. Power Electronics, 2025, 59(4): 58-61. DOI: 10.20222/j.cnki.cn61-1124/tm.2025.04.023.
[3]	张明杰, 杨凯, 刘振, 等. 钠离子电池热安全性研究进展[J]. 电池, 2025, 55(2): 368-375. DOI: 10.19535/j.1001-1579.2025.02.028.
	ZHANG M J, YANG K, LIU Z, et al. Research progress in thermal safety for sodium-ion battery[J]. Battery Bimonthly, 2025, 55(2): 368-375. DOI: 10.19535/j.1001-1579.2025.02.028.
[4]	方永进, 陈重学, 艾新平, 等. 钠离子电池正极材料研究进展[J]. 物理化学学报, 2017, 33(1): 211-241.
	FANG Y J, CHEN Z X, AI X P, et al. Recent developments in cathode materials for Na ion batteries[J]. Acta Physico-Chimica Sinica, 2017, 33(1): 211-241.
[5]	储玉喜,马畅,陈红光,等. 180Ah钠离子电池热失控与产气特性分析[J].储能科学与技术, 2025, 14 (9): 3611-3618. DOI: 10.19799/j.cnki.2095-4239.2025.0242.
	CHU Yuxi, MA Chang, CHEN Hongguang, et al. Thermal runaway and gas production characteristics of a 180 Ah sodium-ion battery[J]. Energy Storage Science and Technology, 2025, 14(9): 3611-3618. DOI: 10.19799/j.cnki.2095-4239.2025.0242.
[6]	李勇琦,李志远, 闻有为, 等. 大容量钠离子电池热失控特性实验研究[J].储能科学与技术, 2025, 14 (4): 1657-1667. DOI: 10.19799/j.cnki.2095-4239.2024.1044.
	LI Yongqi, LI Zhiyuan, WEN Youwei, et al. Experimental study of thermal runaway characteristics of large-capacity sodium-ion batteries[J]. Energy Storage Science and Technology, 2025, 14(4): 1657-1667.
[7]	刘青宜. 钠离子电池的储能机制与性能提升策略[J]. 储能科学与技术, 2024, 13(6): 1871-1873. DOI: 10.19799/j.cnki.2095-4239. 0528.
	LIU Q Y. Energy storage mechanism and performance enhancement strategies of sodium-ion batteries[J]. Energy Storage Science and Technology, 2024, 13(6): 1871-1873. DOI: 10.19799/j.cnki. 2095-4239.0528.
[8]	GUI Q H, XU B, YU K, et al. Comparison of NaNi_1/3Fe_1/3Mn_1/3O₂ and Na₄Fe₃(PO4)₂(P₂O₇) cathode sodium-ion battery behavior under overcharging induced thermal runaway[J]. Chemical Engineering Journal, 2024, 497: 154732. DOI: 10.1016/j.cej. 2024.154732.
[9]	PALANISAMY M, REDDY BODDU V R, SHIRAGE P M, et al. Discharge state of layered P2-type cathode reveals unsafe than charge condition in thermal runaway event for sodium-ion batteries[J]. ACS Applied Materials & Interfaces, 2021, 13(27): 31594-31604. DOI: 10.1021/acsami.1c04482.
[10]	DU W Y, CHEN J L, XING Z X, et al. Battery fault diagnosis and thermal runaway warning based on the feature-exponential-function and dynamic time warping method[J]. Journal of Energy Storage, 2023, 72: 108236. DOI: 10.1016/j.est. 2023. 108236.
[11]	LIU Y P, YANG L J, LIAO R J, et al. Data-driven internal temperature estimation methods for sodium-ion battery using electrochemical impedance spectroscopy[J]. Journal of Energy Storage, 2024, 87: 111426. DOI: 10.1016/j.est.2024.111426.
[12]	WANG J F, CHEN B W, LI Y H, et al. Multiparameter warning of lithium-ion battery overcharge-thermal runaway[J]. Journal of Energy Storage, 2024, 78: 110088. DOI: 10.1016/j.est. 2023. 110088.
[13]	WEI T, XIAN X L, DOU S X, et al. Comprehensive analysis and mitigation strategies for safety issues of sodium-ion batteries[J]. Rare Metals, 2024, 43(4): 1343-1349. DOI: 10.1007/s12598-023-02347-4.
[14]	曹志成, 管敏渊, 楼平, 等. 锂离子电池热失控早期特征及预警方法综述[J]. 电源技术, 2024, 48(5): 771-780.
	CAO Z C, GUAN M Y, LOU P, et al. Review on early thermal runaway characteristic signal and warning method of lithium ion battery[J]. Chinese Journal of Power Sources, 2024, 48(5): 771-780.
[15]	李铜, 罗军, 赵丽亚, 等. 大容量锂电池安全阀应力检测及热失控预警[J/OL]. 电力电子技术, 2025: 1-7. (2025-06-11). https://link.cnki.net/doi/10.20222/j.cnki.cn61-1124/tm.20250611.002.
	LI T, LUO J, ZHAO L Y, et al. Stress monitoring and thermal runaway early warning for safety valves in large-capacity lithium-ion batteries[J/OL]. Power Electronics, 2025: 1-7. (2025-06-11). https://link.cnki.net/doi/10.20222/j.cnki.cn61-1124/tm. 20250611.002.
[16]	马敬轩, 赖铱麟, 吕娜伟, 等. 锂离子电池智能传感监测及预警技术[J]. 电工技术学报, 2025, 40(3): 941-963. DOI: 10.19595/j.cnki. 1000-6753.tces.232137.
	MA J X, LAI Y L, LÜ N W, et al. Lithium-ion battery intelligent sensing monitoring and early warning technology[J]. Transactions of China Electrotechnical Society, 2025, 40(3): 941-963. DOI: 10.19595/j.cnki.1000-6753.tces.232137.
[17]	陈勇, 张建宇, 李全忠. 多源信息融合的磷酸铁锂电池组过充热失控预警方法研究[J]. 电工技术, 2025(7): 155-158. DOI: 10.19768/j.cnki.dgjs.2025.07.038.
	CHEN Y, ZHANG J Y, LI Q Z. Research on warning method of overshoot heat runaway of lithium iron phosphate battery pack based on multi-source information fusion[J]. Electric Engineering, 2025(7): 155-158. DOI: 10.19768/j.cnki.dgjs.2025.07.038.
[18]	田爱娜, 潘壮壮, 吴铁洲, 等. 基于单频阻抗的锂离子电池热失控分级预警[J]. 电池, 2024, 54(2): 194-199. DOI:10.19535/j.1001-1579.2024.02.011.
	TIAN A N, PAN Z Z, WU T Z, et al. Thermal runaway graded warning of Li-ion battery based on single-frequency impedance[J]. Dianchi(Battery Bimonthly), 2024, 54(2): 194-199. DOI:10.19535/j.1001-1579.2024.02.011.

参数名称	标称容量	额定电压	电池重量	能量密度	交流内阻
数值或范围	185 Ah	3.0 V	4.9 kg±0.2 kg	≥110 Wh/kg	≤0.25 mΩ
参数名称	直流内阻	电池尺寸	充电截止电压	放电截止电压
数值或范围	≤0.8 mΩ	厚72 mm±0.8 mm×宽173.7 mm±0.8 mm×高207 mm±0.8 mm	3.85 V	2.0 V

指标编号	指标名称	说明
A1	电压变化速率(dV/dt)	表征电池电压异常波动情况
A2	温度增长速率(dT/dt)	表征内部副反应强弱
A3	膨胀力增长速率(dP/dt)	表征气体生成速率与膨胀趋势

时间段/min	特征表现	演化阶段
0~22	3个参数的变化速率均较小，结构/电化学系统稳定	第一阶段(早期)
22~56	内部变化变缓	第二阶段(风险堆积)
56~87	膨胀速率显著升高，温度/电压尚未突变	第三阶段(中期过热)
87~100+	膨胀速率剧烈波动，温升开始加快，电压逐步波动	第四阶段(失控前夕)
104	温度急剧上升	发生热失控

阶段	对比项目	单参量预警方法(膨胀力)	基于多参量固定阈值的预警方法	本文所提方法
第一阶段	实际起点/min	0	0	0
	算法识别起点/min	无法识别	1	1
	Δt(误差)	—	+1 min	+1 min

第二阶段	实际起点/min	22	22	22
	算法识别起点/min	46	49	23
	Δt(误差)	+24 min	+27 min	+1 min

第三阶段	实际起点/min	56	56	56
	算法识别起点/min	56	60	62
	Δt(误差)	+0 min	4 min	+2 min

第四阶段	实际起点/min	87	87	87
	算法识别起点/min	97	100	92
	Δt(误差)	+10 min	+13 min	+5 min