SOH estimation of lithium-ion batteries based on multi-feature analysis and LSTM-XGBoost model

doi:10.19799/j.cnki.2095-4239.2024.0289

Abstract

Abstract:

Accurate assessment of the state of health (SOH) of lithium-ion batteries is critical to ensure the safe and stable operation of electric vehicles. However, traditional SOH estimation methods face some challenges in effectively extracting health features (HFs) and relying on large amounts of HFs test data. This paper proposes a SOH estimation method for lithium-ion batteries based on multi-feature analysis and the long short-term memory (LSTM)-eXtreme gradient boosting (XGBoost) model. First, to accurately describe the aging mechanism of the battery, a total of 6 HFs are extracted from the battery charging data in three categories: time, energy and IC. Considering a lot of redundant information among the same type of HFs, a feature processing method based on double correlation is presented to screen out the combined HFs that can accurately characterize the trend of battery degradation. Second, a SOH estimation model based on the LSTM-XGBoost is proposed to solve the problem that the traditional SOH estimation model requires a large amount of HFs test data. In this model, the LSTM algorithm is used to predict the HFs data of the number of battery remaining cycles. At the same time, in order to solve the problem of low computational efficiency in HFs prediction using the LSTM model, the LSTM-XGBoost model is developed to estimate battery SOH. The results show that the proposed method can accurately estimate the SOH of lithium batteries under different test data amounts, and the root-mean square error is kept within 1%, which has high estimation accuracy and robustness.

Key words: lithium-ion battery, State of health, Characteristic analysis, long short-term memory neural network, eXtreme gradient boosting

CLC Number:

TM912

Jizhong Lu, Simin Peng, Xiaoyu Li. SOH estimation of lithium-ion batteries based on multi-feature analysis and LSTM-XGBoost model[J]. Energy Storage Science and Technology, doi: 10.19799/j.cnki.2095-4239.2024.0289.

Figures/Tables 12

Fig.1

Fig.2

Fig.3

Table1

Table 2

Fig.4

Table 3

Fig.5

Fig.6

Table 4

Fig.7

Tab.5

References 23

1	YU Q Q, NIE Y W, PENG S M, et al. Evaluation of the safety standards system of power batteries for electric vehicles in China[J]. Applied Energy, 2023,349: 121674.
2	李放,闵永军,张涌.基于大数据的动力锂电池可靠性关键技术研究综述[J].储能科学与技术,2023,12(6):1981-1994.
	LI F, MIN Y J, ZHANG Y. Research review on key reliability technologies of power lithium batteries based on Big Data [J]. Energy Storage Science and Technology,2023,12(6):1981-1994
3	LIN C P, XU J, HOU J Y, et al. A fast data-driven battery capacity estimation method under non-constant current charging and variable temperature[J]. Energy Storage Materials, 2023. 63: 102967.
4	戴国洪,张道涵,彭思敏等.人工智能在动力电池健康状态预估研究综述[J/OL].机械工程学报:1-18[2024-02-11].
	DAI G H, ZHANG D H, PENG S M, et al. Research review of Artificial Intelligence in Power battery health state prediction [J/OL]. Chinese Journal of Mechanical Engineering :1-18[2024-02-11].
5	HUANG H Y, MENG J H,WANG Y H, et al. A comprehensively optimized lithium-ion battery state-of-health estimator based on Local Coulomb Counting Curve[J]. Applied Energy, 2022. 322: 119469.
6	HOU J Y, XV J, LIN C P, et al. State of charge estimation for lithium-ion batteries based on battery model and data-driven fusion method[J]. Energy, 2024. 290: 130056.
7	Shah A, Shah K, Shah C, et al. State of charge, remaining useful life and knee point estimation based on artificial intelligence and Machine learning in lithium-ion EV batteries: A comprehensive review[J]. Renewable Energy Focus, 2022. 42: 146-164.
8	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/OL].高电压技术,1-14[2024-02-11].
	XIONG Q, DI Z G, JI S C. Review of research progress on health state estimation and life prediction of lithium-ion batteries [J/OL]. High Voltage Technology,1-14[2024-02-11].
9	陈清炀,何映晖,余官定等.模型与数据双驱动的锂电池状态精准估计[J].储能科学与技术,2023,12(1):209-217.
	Chen Q Y, He Y H, Yu G D, et al. Accurate state estimation of lithium batteries driven by model and Data [J]. Energy Storage Science and Technology,2023,12(1):209-217.
10	LING M, HU H Z, Chen J J, et al. Online State-of-Health Estimation Method for Lithium-Ion Battery Based on CEEMDAN for Feature Analysis and RBF Neural Network. in IEEE Journal of Emerging and Selected Topics in Power Electronics [J], 2023, 11(1):187-200.
11	ZHANG C L, LUO L J, YANG Z, et al. Battery SOH estimation method based on gradual decreasing current, double correlation analysis and GRU[J]. Green Energy and Intelligent Transportation, 2023,2(5): 100108
12	YANG J F, CAI Y F and MI C, Lithium-ion battery capacity estimation based on battery surface temperature change under constant-current charge scenario[J]. Energy, 2022. 241: 122879.
13	李乐卿,王鹏,孙万洲等.基于锂离子电池容量增量曲线半峰面积的容量在线估计方法[J/OL].电工技术学报,1-9[2024-02-11].
	LI L Q, WANG P, SUN W Z, et al. Online capacity estimation method based on half-peak area of capacity increment curve of lithium-ion battery [J/OL]. Transactions of China Electrotechnical Society,1-9[2024-02-11].
14	MA Y, CE S, GAO J W, et al. A novel method for state of health estimation of lithium-ion batteries based on improved LSTM and health indicators extraction[J]. Energy, 2022. 251: 123973.
15	JI W, CUI X C, MENG J H, et al. Data-driven transfer-stacking based state of health estimation for lithium-ion batteries. IEEE Trans Ind Electron, 2024,71(1):604-614.
16	Li R R, LI W R and Zhang H N. State of Health and Charge Estimation Based on Adaptive Boosting integrated with particle swarm optimization/support vector machine (AdaBoost-PSO-SVM) Model for Lithium-ion Batteries. International Journal of Electrochemical Science[J], 2022. 17(2): 220212.
17	GAO Z H, XIE H C, YANG X B, et al. SOH estimation method for lithium-ion batteries under low temperature conditions with nonlinear correction[J]. Journal of Energy Storage, 2024. 75: 109690.
18	CHEN S Z, LIANG Z K, YUAN H L,et al. Li-ion battery state-of-health estimation based on the combination of statistical and geometric features of the constant-voltage charging stage[J]. Journal of Energy Storage, 2023. 72: 108647.
19	YU Q Q, NIE Y W, LIU S Z, et al., State of health estimation method for lithium-ion batteries based on multiple dynamic operating conditions[J]. Journal of Power Sources, 2023. 582: 233541.
20	PENG S M, SUN Y X, LIU D D, et al. State of health estimation of lithium-ion batteries based on multi-health features extraction and improved long short-term memory neural network[J], Energy, 2023, 282:128956.
21	NA Y, YAO Y B, JIA Z D, et al. Online battery health diagnosis for electric vehicles based on DTW-XGBoost[J]. Energy Reports, 2022. 8: 121-128.
22	S. Jafari and Y. -C. Byun. Prediction of the Battery State Using the Digital Twin Framework Based on the Battery Management System[J]. IEEE Access, 2022,10:124685-124696..
23	JI W, Fang L C, GUANG Z D, et al. State of health estimation of lithium-ion battery with improved radial basis function neural network[J]. Energy, 2023, 262:125380.

特征		灰色关联度
特征		B0005	B0006	B0007
时间特征	HF1	0.9271	0.9096	0.9443
时间特征	HF2	0.8080	0.7867	0.7866
能量特征	HF3	0.9106	0.9601	0.9411
能量特征	HF4	0.8235	0.7550	0.7943
IC曲线特征	HF5	0.9500	0.9409	0.9259
IC曲线特征	HF6	0.8878	0.8512	0.7729

电池	特征	主要HFs		灰色关联度
B0005	HF₁	HF₅	0.9817	HF₁
	HF₂		0.7763	HF₁
	HF₃		0.9513	HF₃
	HF₄		0.7892	HF₃
B0006	HF₁	HF₃	0.9208	HF₁
	HF₂		0.7838	HF₁
	HF₅		0.9657	HF₅
	HF₆		0.8593	HF₅
B0007	HF₃	HF₁	0.9814	HF₃
	HF₄		0.8114	HF₃
	HF₅		0.9715	HF₅
	HF₆		0.7948	HF₅

电池	特征	MAE (%)	MAPE (%)	RMSE (%)
B0005	HF₅	0.2184	1.1698	0.3386
	HF₁	0.5214	1.7601	0.6033
	HF₃	0.6072	1.9265	0.7928
B0006	HF₃	0.4523	2.2077	0.6111
	HF₁	0.3546	2.2734	0.6049
	HF₅	0.4164	1.4824	0.8212
B0007	HF₁	0.2247	0.8649	0.4459
	HF₃	0.3771	1.1927	0.4801
	HF₅	0.4371	1.7686	0.5080

电池	预测起点	MAE （%）	MAPE（%）	RMSE（%）
B0005	80	0.5621	0.7936	0. 8677
	100	0.3826	0.5609	0.4585
	120	0.3246	0.4868	0. 3638
B0006	80	0. 9228	1.3509	1.1205
	100	0.6338	0.9663	0.7988
	120	0.6220	0.9928	0.8434
B0007	80	0.4201	0.5475	0.7471
	100	0.3868	0.5277	0.5044
	120	0.3526	0.4851	0.4395

电池	模型	MAE （%）	MAPE （%）	RMSE （%）	计算耗时(s)
B0005	RBF	0.8735	1.8355	1.3354	2.87
	SVM	0.6938	1.0219	0.7567	2.45
	LSTM	0.5733	0.8433	0.6317	12.78
	本文模型	0.3826	0.5609	0.4585	4.30
B0006	RBF	1.0926	1.5557	1.2837	2.93
	SVM	1.2394	1.9945	1.6654	2.52
	LSTM	0.7645	1.1567	0.9451	13.67
	本文模型	0.6338	0.9663	0.7988	4.41
B0007	RBF	0.7312	1.0245	0.8934	3.17
	SVM	0.7155	0.9761	0.8099	2.98
	LSTM	0.6477	0.8829	0.7364	13.04
	本文模型	0.3868	0.5277	0.5044	4.24