基于多尺度健康因子-BEAST分解和SARIMA模型结合的锂离子电池剩余使用寿命预测

Remaining Useful Life Prediction of Lithium-Ion Batteries Based on Combination of Multiscale Health Indicators-BEAST Decomposition and SARIMA Modeling

摘要

锂离子电池的剩余使用寿命(RUL)预测对电池管理和安全性至关重要．现有的RUL预测方法多依赖大量历史数据，且在复杂工况下精度较低，计算负担重．为解决这些问题，本文结合健康因子(HI)、贝叶斯时序分解估计器(BEAST)和季节性差分自回归移动平均模型(SARIMA)，提出了一种新颖的RUL预测方法．与传统方法不同，本文创新性地采用HI替代最大可放电容量，能够更精确地反映电池衰退过程；同时，结合贝叶斯时序分解估计器对HI进行分解与重构，提高了预测精度，减少了对大量历史数据的依赖；最后，利用季节性差分自回归移动平均模型对电池衰退的时序数据进行建模，显著提高了预测精度和计算效率．实验结果表明，以动态工况电池(CS#7)为例，所提方法在电池衰减5%时，最大相对误差小于2%，衰减10%时小于4.31%；相比LSTM和LSSVM方法，本文方法在MAE上分别降低了16.6%和25.9%，计算效率分别提高了55.2%和22.8%．

Abstract

Predicting the remaining useful life(RUL)of lithium-ion batteries is critical for battery management and safety. Existing RUL prediction methods mostly rely on a large amount of historical data and are less accurate and computationally burdensome under complex operating conditions. To address these issues, this paper proposes a novel RUL prediction method by combining the health indicators(HI), the Bayesian estimator of abrupt change, seasonality, and trend(BEAST), and the seasonal autoregressive integrated moving average(SARIMA)model. Different from the traditional method, this paper innovatively adopts HI to replace the maximum dischargeable capacity, which can reflect the battery decline process more accurately. Meanwhile, the decomposition and reconstruction of HI in combination with BEAST improves the prediction accuracy and reduces the dependence on a large amount of historical data. Finally, the SARIMA model for the time-series data of battery decline considerably improves the prediction accuracy and computational efficiency. The experimental results indicate that for batteries under dynamic operating conditions(CS#7), the proposed method maintains maximum relative errors below 2% at 5% capacity degradation and 4.31% at 10% degradation. Compared with the long short-term memory and least square support vector machine methods, the method in this paper reduces the mean absolute error by 16.6% and 25.9% and improves the computational efficiency by 55.2% and 22.8%, respectively.

关键词

锂离子电池 / 剩余使用寿命 / 健康因子 / BEAST分解 / SARIMA模型

Key words

lithium-ion batteries / remaining useful life / health indicator(HI) / BEAST decomposition / seasonal autoregressive integrated moving average(SARIMA) model

参考文献

[1] Yao F, He W X, Wu Y X, et al. Remaining useful life prediction of lithium—ion batteries using a hybrid model[J]. Energy, 2022, 248:123622.

[2] 姚芳, 张楠, 黄凯. 锂离子电池状态估算与寿命预测综述[J]. 电源学报, 2020, 18(3):175-183.

[3] Yao Fang, Zhang Nan, Huang Kai. Review of state estimation and life prediction for lithiumion batteries[J]. Journal of Power Supply, 2020, 18(3):175-183(in Chinese).

[4] Li M, Xu W T, Zhang S W, et al. State of health estimation and battery management:A review of health indicators, models and machine learning[J]. Materials, 2025, 18(1): 145— 145.

[5] Fei Z C, Yang F F, Tsui K L, et al. Early prediction of battery lifetime via a machine learning based framework[J]. Energy, 2021, 225:120205.

[6] 戴俊彦, 夏明超, 陈奇芳. 基于双重注意力机制的电池SOH估计和RUL预测编解码模型[J]. 电力系统自动化, 2023, 47(6):168-177.

[7] Dai Junyan, Xia Mingchao, Chen Qifang. Encoding and decoding model of state of health estimation and remaining useful life prediction for batteries based on dual—stage attention mechanism[J]. Automation of Electric Power Systems, 2023, 47(6):168-177(in Chinese).

[8] Wu L F, Fu X H, Guan Y. Review of the remaining useful life prognostics of vehicle lithium—ion batteries using data—driven methodologies[J]. Applied Sciences, 2016, 6(6):166.

[9] Lui Y H, Li M, Downey A, et al. Physics—based prognostics of implantable—grade lithium—ion battery for remaining useful life prediction[J]. Journal of Power Sources, 2021, 485:229327.

[10] 戴海峰, 张艳伟, 魏学哲, 等. 锂离子电池剩余寿命预测研究[J]. 电源技术, 2019, 43(12):2029-2035.

[11] Dai Haifeng, Zhang Yanwei, Wei Xuezhe, et al. Review of remaining useful life prediction for lithium ion battery[J]. Chinese Journal of Power Sources, 2019, 43(12):2029-2035(in Chinese).

[12] 范立明, 王崑声, 钱诚. 故障物理与粒子滤波融合的锂离子电池寿命预测方法[J]. 兵器装备工程学报, 2020, 41(9):171-175.

[13] Fan Liming, Wang Kunsheng, Qian Cheng. Remaining useful life prediction of lithium battery based on physics of failure and particle filtering[J]. Journal of Ordnance Equipment Engineering, 2020, 41(9):171-175(in Chinese).

[14] 熊庆, 邸振国, 汲胜昌. 锂离子电池健康状态估计及寿命预测研究进展综述[J]. 高电压技术, 2024, 50(3):1182-1195.

[15] Xiong Qing, Di Zhenguo, Ji Shengchang. Review on health state estimation and life prediction of lithium—ion batteries[J]. High Voltage Engineering, 2024, 50(3):1182-1195(in Chinese).

[16] 孙冬, 许爽. 梯次利用锂离子电池健康状态预测[J]. 电工技术学报, 2018, 33(9):2121-2129.

[17] Sun Dong, Xu Shuang. State of health prediction of second—use lithium—ion battery[J]. Transactions of China Electrotechnical Society, 2018, 33(9):2121-2129(in Chinese).

[18] Sultan A Y, Eladl A A, Hassan A M, et al. Enhancing electric vehicle battery lifespan:Integrating active balancing and machine learning for precise RUL estimation[J]. Scientific Reports, 2025, 15(1):777.

[19] Ahwiadi M, Wang W. Battery health monitoring and remaining useful life prediction techniques:A review of technologies[J]. Batteries, 2025, 11(1):31.

[20] Li L, Saldivar A A F, Bai Y, et al. Battery remaining useful life prediction with inheritance particle filtering[J]. Energies, 2019, 12(14):2784.

[21] 吴诗淼, 王文波, 朱婷, 等. 基于跳跃连接多尺度CNN的锂离子电池剩余寿命预测[J]. 太阳能学报, 2024, 45(7):199-208.

[22] Wu Shimiao, Wang Wenbo, Zhu Ting, et al. Prediction of residual life of lithium ion battery based on multiscale CNN with jump connection[J]. Acta Energiae Solaris Sinica, 2024, 45(7):199-208(in Chinese).

[23] Chen Z W, Shi N, Ji Y F, et al. Lithium—ion batteries remaining useful life prediction based on BLS—RVM[J]. Energy, 2021, 234:121269.

[24] Chen L G H, Zhang Y, Zheng Y, et al. Remaining useful life prediction of lithium—ion battery with optimal input sequence selection and error compensation[J]. Neurocomputing, 2020, 414:245-254.

[25] Wang S L, Fernandez C, Cao W, et al. An adaptive working state iterative calculation method of the power battery by using the improved Kalman filtering algorithm and considering the relaxation effect[J]. Journal of Power Sources, 2019, 428:67-75.

[26] Pang X Q, Liu X Y, Jia J F, et al. A lithium—ion batery remaining useful life prediction method based on the incremental capacity analysis and Gaussian process regression[J]. Microelectronics Reliability, 2021, 127:114405.

[27] Liu L, Sun W, Yue C X, et al. Remaining useful life estimation of lithium—ion batteries based on small sample models[J]. Energies, 2024, 17(19):4932.

[28] 方斯顿, 刘龙真, 孔赖强. 基于双向长短期记忆网络含间接健康指标的锂电池SOH估计[J]. 电力系统自动化, 2024, 48(4):160-168.

[29] Fang Sidun, Liu Longzhen, Kong Laiqiang. State—of—health estimation for lithium—ion batteries incorporating indirect health indicators based on bi—directional long short—term memory networks[J]. Automation of Electric Power Systems, 2024, 48(4):160-168(in Chinese).

[30] Saleem U, Liu W J, Riaz S, et al. TransRUL:A transformer—based multihead attention model for enhanced prediction of battery remaining useful life[J]. Energies, 2024, 17(16):3976.

[31] Xie Y X, Wang S L, Zhang G X, et al. A review of data—driven whole—life state of health prediction for lithium—ion batteries:Data preprocessing, aging characteristics, algorithms, and future challenges[J]. Journal of Energy Chemistry, 2024, 97:630-649.

[32] Alipour M, Ziebert C, Conte V F, et al. A review on temperature—dependent electrochemical properties, aging, and performance of lithium—ion cells[J]. Batteries, 2020, 6(3):35.

[33] Wang Z, Zhao Q J, Wang S J, et al. Aging and post—aging thermal safety of lithium—ion batteries under complex operating conditions:A comprehensive review[J]. Journal of Power Sources, 2024, 623:235453.

[34] Lyu R F, Pang J L, Zhang J M, et al. The impacts of disturbances on mountain ecosystem services:Insights from BEAST and Bayesian network[J]. Applied Geography, 2024, 162:103143.

[35] Hu Y F, Wang S L, Huang J H, et al. A novel seasonal autoregressive integrated moving average method for the accurate lithium—ion battery residual life prediction[J]. International Journal of Electrochemical Science, 2022, 17(5):220555.

[36] Zhang W Q, Lin Z, Liu X L. Short—term offshore wind power forecasting—A hybrid model based on discrete wavelet transform(DWT), seasonal autoregressive integrated moving average(SARIMA), and deep—learning—based long short—term memory(LSTM)[J]. Renewable Energy, 2022, 185:611-628.

[37] Sayah M, Guebli D, Al Masry Z, et al. Robustness testing framework for RUL prediction deep LSTM networks[J]. ISA Transactions, 2021, 113:28-38.

[38] Li X, Ma Y, Zhu J J. An online dual filters RUL prediction method of lithium—ion battery based on unscented particle filter and least squares support vector machine[J]. Measurement, 2021, 184:109935.

基金资助

河北省自然科学基金资助项目(E2022202056)