基于DAUKF-LSTM协同的锂离子电池储能系统低温加热目标温度动态优化

doi:10.19595/j.cnki.1000-6753.tces.251357

Abstract
Figure/Table
References
Related Citation (10)

Download: PDF (6445 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract The performance of lithium-ion battery energy storage systems degrades severely in low- temperature environments. Electrochemical reactions slow down, internal resistance increases, and available capacity drops, posing a major challenge for applications in cold climates. A critical yet often overlooked factor is the strategy for setting the heating target temperature. Conventional approaches typically rely on a fixed temperature value, determined by general engineering experience and maintained throughout the discharge cycle. Such a static method fails to account for the dynamic changes in the battery's internal states during operation. Consequently, energy utilization is suboptimal, excessive heating energy is consumed, and overall system efficiency is compromised.
This study introduces a dynamic optimization method for the low-temperature heating target with a dual adaptive unscented Kalman filter (DAUKF) and a long short-term memory (LSTM) neural network. An electro-thermal coupling model is constructed to capture the intricate interactions between electrical and thermal behaviors. Since the state of charge (SOC) and state of temperature (SOT) are highly nonlinear and tightly coupled parameters, a DAUKF-based joint estimation framework is proposed. Compared to the conventional dual extended Kalman filter (DEKF), the DAUKF demonstrates superior accuracy and robustness under strongly nonlinear, time-varying conditions, effectively mitigating error accumulation and state drift.
Then, an efficiency-temperature mapping model is constructed using an LSTM network to predict the optimal discharge temperature for each SOC level. It overcomes the limitations of fixed empirical settings by learning the complex, non-linear relationship between SOC, temperature, and discharge efficiency from extensive experimental data. The LSTM captures the battery's efficiency dynamics during low-temperature discharge, enabling real-time prediction of the optimal heating target. Its strong generalization capability allows it to adapt to varying operational scenarios.
Finally, a closed-loop dynamic control framework is constructed that operates cyclically through three core stages: state estimation, temperature prediction, and feedback-based optimization. The DAUKF continuously provides real-time SOC and SOT estimates that are fed into the LSTM model, which then generates the optimal heating target temperature in real time. This process enables multi-stage, full-process dynamic temperature regulation, where the heating target adjusts intelligently based on the battery's actual condition. As a result, the battery is maintained at the most energy-efficient temperature at any given time, enhancing discharge perfor- mance while minimizing heating energy consumption.
Experimental validation was conducted under various low-temperature discharge scenarios. The results confirm that the proposed DAUKF-LSTM collaborative optimization method significantly improves battery discharge efficiency. The average energy-efficiency gain consistently outperforms that of traditional constant- temperature control strategies. This paper contributes a novel and effective paradigm for intelligent thermal management in cold-region energy storage applications. By enabling dynamic, self-adaptive temperature control, it paves the way for more efficient and reliable next-generation smart energy systems.

Key words： Low-temperature lithium-ion battery energy storage systems SOC-SOT joint estimation dual adaptive unscented Kalman filter (DAUKF) long short-term memory (LSTM) network dynamic optimization of heating target temperature

Received: 03 August 2025

PACS:

TM912

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Liu Peijin
	Sun Wenchi
	Yao Qian

Cite this article:

Liu Peijin,Sun Wenchi,Yao Qian. Dynamic Optimization of Low-Temperature Heating Target for Energy Storage Systems Based on DAUKF-LSTM Estimation[J]. Transactions of China Electrotechnical Society, 2026, 41(14): 4974-4987.

URL:

https://dgjsxb.ces-transaction.com/EN/10.19595/j.cnki.1000-6753.tces.251357 OR https://dgjsxb.ces-transaction.com/EN/Y2026/V41/I14/4974

[1] Mei Pan, Zhang Yuan, Zhang Wei.Low-temperature lithium-ion batteries: challenges and progress of surface/interface modifications for advanced perfor-mance[J]. Nanoscale, 2023, 15(3): 987-997.
[2] Belgibayeva A, Rakhmetova A, Rakhatkyzy M, et al.Lithium-ion batteries for low-temperature appli-cations: Limiting factors and solutions[J]. Journal of Power Sources, 2023, 557: 232550.
[3] 何锡添, 孙丙香, 阮海军, 等. 锂离子电池变频变幅交流低温自加热策略[J]. 电工技术学报, 2019, 34(9): 1798-1805.
He Xitian, Sun Bingxiang, Ruan Haijun, et al.A variable-frequency and variable-amplitude AC low-temperature self-heating strategy for lithium-ion battery[J]. Transactions of China Electrotechnical Society, 2019, 34(9): 1798-1805.
[4] 及非凡, 陈子豪, 李艳君. 基于电驱系统高频注入的电动汽车电池包快速加热控制算法[J]. 电工技术学报, 2024, 39(20): 6316-6327.
Ji Feifan, Chen Zihao, Li Yanjun.Motor drive system based battery self-heating for electric vehicles using high frequency injection frame[J]. Transactions of China Electrotechnical Society, 2024, 39(20): 6316-6327.
[5] 韦鹏, 舒小杨, 朱文超, 等. 基于长短期记忆网络的动力电池组温度场时空建模[J]. 电工技术学报, 2025, 40(13): 4306-4315.
Wei Peng, Shu Xiaoyang, Zhu Wenchao, et al.Spatio-temporal modeling of temperature field for power battery packs based on long short-term memory network[J]. Transactions of China Electrotechnical Society, 2025, 40(13): 4306-4315.
[6] 王新栋, 董政, 王书华, 等. 基于改进开路电压模型和自适应平方根无迹卡尔曼滤波的锂离子电池宽温度多工况 SOC 估计[J]. 电工技术学报, 2024, 39(24): 7950-7964.
Wang Xindong, Dong Zheng, Wang Shuhua, et al.State-of-charge estimation for lithium-ion batteries across wide temperature range and multiple working conditions based on improved open-circuit voltage model and adaptive square root unscented Kalman filter algorithm[J]. Transactions of China Electro-technical Society, 2024, 39(24): 7950-7964.
[7] Rahmani A, Dibaj M, Akrami M.Recent advan-cements in battery thermal management systems for enhanced performance of Li-ion batteries: a com-prehensive review[J]. Batteries, 2024, 10(8): 265.
[8] Ma Shuai, Jiang Modi, Tao Peng, et al.Temperature effect and thermal impact in lithium-ion batteries: a review[J]. Progress in Natural Science: Materials International, 2018, 28(6): 653-666.
[9] Ji Changwei, Liu Dianqing, Liu Yangyi, et al.Effect of low temperature and high-rate cyclic aging on thermal characteristics and safety of lithium-ion batteries[J]. Process Safety and Environmental Pro-tection, 2024, 188: 1514-1526.
[10] 郭宝贵, 马潇男. 基于分数阶模型的锂离子电池SOC估计[J]. 电池, 2024, 54(5): 634-639.
Guo Baogui, Ma Xiaonan.SOC estimation of Li-ion battery based on fractional order modeling[J]. Battery Bimonthly, 2024, 54(5): 634-639.
[11] Yu Chao, Zhu Jiangong, Liu Wenxue, et al.Co-estimation of state-of-charge and state-of-temperature for large-format lithium-ion batteries based on a novel electrothermal model[J]. Green Energy and Intelligent Transportation, 2024, 3(4): 100152.
[12] Ding Linwang, Wen Chenglin.High-order extended Kalman filter for state estimation of nonlinear systems[J]. Symmetry, 2024, 16(5): 617.
[13] Ding Jie, Chen Lijuan, Cao Zhengxin, et al.Con-vergence analysis of the modified adaptive extended Kalman filter for the parameter estimation of a brushless DC motor[J]. International Journal of Robust and Nonlinear Control, 2021, 31(16): 7606-7620.
[14] 庞松岩, 肖传亮, 彭克, 等. 基于轻量级梯度提升机的锂离子电池状态联合估计[J]. 电力系统自动化, 2025, 49(20): 200-207.
Pang Songyan, Xiao Chuanliang, Peng Ke, et al.Joint estimation of lithium-ion battery state based on light gradient boosting machine[J]. Automation of Electric Power Systems, 2025, 49(20): 200-207.
[15] Capasso C, Castiglione T, Perrone D, et al.Thermal management of Li-ion batteries: Experimentally validated model of active temperature control strategies with liquid cooling and multi-objective optimization[J]. Journal of Energy Storage, 2025, 128: 117254.
[16] Wu Yue, Huang Zhiwu, Li Dongjun, et al.Optimal battery thermal management for electric vehicles with battery degradation minimization[J]. Applied Energy, 2024, 353: 122090.
[17] Chen Jiahui, Xuan Dongji, Chen Cong, et al.Structure optimization of air-cooled battery thermal management system based on neural network[J]. Ionics, 2023, 29(7): 2773-2782.
[18] Huang Dan, Huang Shuang.Optimization design and thermodynamic analysis of thermal management system for new energy vehicle power batteries[J]. International Journal of Heat and Technology, 2024, 42(1): 253-262.
[19] Luo Mingyun, Ling Ziye, Zhang Zhengguo, et al.A fast-response preheating system coupled with supercapacitor and electric conductive phase change materials for lithium-ion battery energy storage system at low temperatures[J]. Journal of Energy Storage, 2023, 73: 109255.
[20] Han Shaojian, Li Chengwei, Ding Jifeng, et al.An improved PSO-based DC discharge heating strategy for lithium-ion batteries at low temperatures[J]. Energies, 2025, 18(9): 2261.
[21] Tao Zhenyi, Lin Cheng, Tian Yu, et al.A systematic approach for determining the optimal battery preheating cut-off temperature for electric vehicles operating in cold climates[J]. Applied Energy, 2025, 383: 125308.
[22] 姬鹏, 吕泽旭. 考虑温度及老化的储能用锂离子电池组荷电状态估算算法[J]. 电工技术学报, 2025, 40(17): 5667-5682.
Ji Peng, Lü Zexu.SOC estimation algorithm of lithium-ion battery pack for energy storage con-sidering temperature and aging[J]. Transactions of China Electrotechnical Society, 2025, 40(17): 5667-5682.
[23] 常小兵, 侯宗尚, 刘连起, 等. 基于电热耦合效应的锂电池荷电状态与温度状态联合估计[J]. 储能科学与技术, 2024, 13(4): 1142-1153.
Chang Xiaobing, Hou Zongshang, Liu Lianqi, et al.Joint estimation of the state of charge and temperature of lithium batteries based on the electric thermal coupling effect[J]. Energy Storage Science and Technology, 2024, 13(4): 1142-1153.
[24] Cao Yuan, Chen Long, Wang Chunsheng.Coor-dinated thermal and electrical balancing for lithium-ion cells[J]. Energies, 2025, 18(16): 4231.
[25] Shreasth, Ooi C A, Khan N, et al. A novel active lithium-ion cell balancing method based on charging and discharging state of power in electric vehicles[J]. Scientific Reports, 2025, 15: 15764.
[26] Huang Congsheng, Cheng Zheyuan, Chow M Y.A robust and efficient state-of-charge estimation methodology for serial-connected battery packs: most significant cell methodology[J]. IEEE Access, 2021, 9: 74360-74369.
[27] 陈智伟, 张婷婷, 张维戈, 等. 电池组不一致性参数估算方法[J/OL]. 电源学报, 1-13 [2025-10-31]. http://link.cnki.net/urlid/12.1420.TM.20250513.1838.002.
Chen Zhiwei, Zhang Tingting, Zhang Weige, et al. Battery parameter estimation method based on simu-lation database[J/OL]. Journal of Power Supply, 1-13 [2025-10-31]. http://link.cnki.net/urlid/12.1420.TM.20250513.1838.002.
[28] Jia Fu, Lee G.Simulation of battery thermal management system for large maritime electric ship's battery pack[J]. Energies, 2024, 17(18): 4587.
[29] Rasool G, Wang Xinhua, Tao Sun, et al.Recent advancements in battery thermal management system (BTMS): a review of performance enhancement techniques with an emphasis on nano-enhanced phase change materials[J]. Heliyon, 2024, 10(17): e36950.
[30] Csomós B, Kocsis Szürke S, Fodor D.Comparison of coupled electrochemical and thermal modelling strategies of 18650 Li-ion batteries in finite element analysis: a review[J]. Materials, 2023, 16(24): 7613.
[31] 王帮亭, 王莉, 郑聪, 等. 锂电池联合状态估计与故障诊断方法研究[J]. 电气技术, 2025, 26(7): 1-12.
Wang Bangting, Wang Li, Zheng Cong, et al.Research on joint state estimation and fault diagnosis method for lithium battery[J]. Electrical Engineering, 2025, 26(7): 1-12.