RUL prediction of lithium-ion battery based on differential voltage and Elman neural network

doi:10.19799/j.cnki.2095-4239.2021.0158

Abstract

Abstract:

The prediction of the Remaining Useful Life (RUL) of lithium-ion battery has significant value for the study of the usage and maintenance of batteries. A method for RUL prediction of lithium-ion batteries based on differential voltage and Elman neural network is proposed. Firstly, based on the National Aeronautics and Space Administration' (NASA) lithium-ion battery data set, the differential voltage curve and charge-discharge curve of the battery are analyzed, and the characteristic quantity of battery capacity degradation is extracted. Secondly, the correlation between characteristic parameters is studied by Pearson method. The inflexion point of charging differential voltage curve, peak value of discharging differential voltage curve, as well as the discharging time and resting time are determined as indirect health factors of battery RUL prediction. Finally, an Elman neural network employed the inputs of above indirect health factors and the output of battery capacity is established to predict RUL of lithium-ion batteries. The comparative experiments of four kinds of battery capacity prediction based on different indirect health factors and different neural networks show that the prediction accuracy of battery life can be improved by adding the initial inflection point of charging differential voltage curve and the peak value of discharge differential voltage curve into the indirect health factors, and the Elman neural network can accurately predict the battery capacity. The mean root mean square error (RMSE) of the prediction of battery RUL based on different cycle times is 1.55%.

Key words: lithium-ion batteries, RUL, differential voltage, Elman neural network, correlation coefficient

CLC Number:

TM 911.3

Lianbing LI, Sijia LI, Jie LI, Kun SUN, Zhengping WANG, Haiyue YANG, Bing GAO, Shaobo YANG. RUL prediction of lithium-ion battery based on differential voltage and Elman neural network[J]. Energy Storage Science and Technology, 2021, 10(6): 2373-2384.

Figures/Tables 16

Table 1

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Table 2

Fig. 5

Table 3

Fig. 6

Fig. 7

Table 4

Fig. 8

Table 5

Fig. 9

Fig. 10

Fig. 11

References 22

1	LI T M, PEI H, PANG Z N, et al. A sequential Bayesian updated Wiener process model for remaining useful life prediction[J]. IEEE Access, 2019, 8: 5471-5480.
2	KHODADADI SADABADI K, JIN X, RIZZONI G. Prediction of remaining useful life for a composite electrode lithium ion battery cell using an electrochemical model to estimate the state of health[J]. Journal of Power Sources, 2021, 481: doi: 10.1016/j.jpowsour. 2020.228861.
3	GUHA A, PATRA A. Online estimation of the electrochemical impedance spectrum and remaining useful life of lithium-ion batteries[J]. IEEE Transactions on Instrumentation and Measurement, 2018, 67(8): 1836-1849.
4	ZHANG H, MO Z L, WANG J Y, et al. Nonlinear-drifted fractional Brownian motion with multiple hidden state variables for remaining useful life prediction of lithium-ion batteries[J]. IEEE Transactions on Reliability, 2020, 69(2): 768-780.
5	WANG D, YANG F F, ZHAO Y, et al. Prognostics of lithium-ion batteries based on state space modeling with heterogeneous noise variances[J]. Microelectronics Reliability, 2017, 75: 1-8.
6	WANG H Y, SONG W Q, ZIO E, et al. Remaining useful life prediction for lithium-ion batteries using fractional Brownian motion and Fruit-fly optimization algorithm[J]. Measurement, 2020, 161: doi: 10.1016/j.measurement.2020.107904.
7	张婷婷, 于明, 李宾, 等. 基于Wavelet降噪和支持向量机的锂离子电池容量预测研究[J]. 电工技术学报, 2020, 35(14): 3126-3136.
	ZHANG T T, YU M, LI B, et al. Capacity prediction of lithium-ion batteries based on wavelet noise reduction and support vector machine[J]. Transactions of China Electrotechnical Society, 2020, 35(14): 3126-3136.
8	LI P H, ZHANG Z J, XIONG Q Y, et al. State-of-health estimation and remaining useful life prediction for the lithium-ion battery based on a variant long short term memory neural network[J]. Journal of Power Sources, 2020, 459: doi: 10.1016/j.jpowsour. 2020.228069.
9	李练兵, 祝亚尊, 田永嘉, 等. 基于Elman神经网络的锂离子电池RUL间接预测研究[J]. 电源技术, 2019, 43(6): 1027-1031.
	LI L B, ZHU Y Z, TIAN Y J, et al. RUL indirect prediction of lithium-ion battery based on Elman neural network[J]. Chinese Journal of Power Sources, 2019, 43(6): 1027-1031.
10	LIU K L, SHANG Y L, OUYANG Q, et al. A data-driven approach with uncertainty quantification for predicting future capacities and remaining useful life of lithium-ion battery[J]. IEEE Transactions on Industrial Electronics, 2020, 68(4): 3170-3180.
11	REN L, DONG J B, WANG X K, et al. A data-driven auto-CNN-LSTM prediction model for lithium-ion battery remaining useful life[J]. IEEE Transactions on Industrial Informatics, 2021, 17(5): 3478-3487.
12	WANG L M, PAN C F, LIU L, et al. On-board state of health estimation of LiFePO₄ battery pack through differential voltage analysis[J]. Applied Energy, 2016, 168: 465-472.
13	XU T T, PENG Z, WU L F. A novel data-driven method for predicting the circulating capacity of lithium-ion battery under random variable current[J]. Energy, 2021, 218: doi: 10.1016/j.energy.2020.119530.
14	郭永芳, 黄凯, 李志刚. 基于短时搁置端电压压降的快速锂离子电池健康状态预测[J]. 电工技术学报, 2019, 34(19): 3968-3978.
	GUO Y F, HUANG K, LI Z G. Fast state of health prediction of lithium-ion battery based on terminal voltage drop during rest for short time[J]. Transactions of China Electrotechnical Society, 2019, 34(19): 3968-3978.
15	李练兵, 季亮, 祝亚尊, 等. 等效循环电池组剩余使用寿命预测[J]. 工程科学学报, 2020, 42(6): 796-802.
	LI L B, JI L, ZHU Y Z, et al. Investigation of RUL prediction of lithium-ion battery equivalent cycle battery pack[J]. Chinese Journal of Engineering, 2020, 42(6): 796-802.
16	HONKURA K, TAKAHASHI K, HORIBA T. Capacity-fading prediction of lithium-ion batteries based on discharge curves analysis[J]. Journal of Power Sources, 2011, 196(23): 10141-10147.
17	BLOOM I, JANSEN A N, ABRAHAM D P, et al. Differential voltage analyses of high-power, lithium-ion cells: 1. Technique and application[J]. Journal of Power Sources, 2005, 139(1/2): 295-303.
18	SAHA B, GOEBEL K. Battery data set[R]. NASA Ames Prognostics Data Repository, 2007.
19	郝雪玲. 锂离子电池健康状态多指标融合和剩余寿命预测方法研究[D]. 哈尔滨: 哈尔滨理工大学, 2019.
	HAO X L. Study on multi-health indicators fusion and remaining useful life prediction for lithium-ion batteries[D]. Harbin: Harbin University of Science and Technology, 2019.
20	ZHANG S Z, ZHAI B Y, GUO X, et al. Synchronous estimation of state of health and remaining useful lifetime for lithium-ion battery using the incremental capacity and artificial neural networks[J]. Journal of Energy Storage, 2019, 26: doi: 10.1016/j.est.2019.100951.
21	LI X Y, YUAN C G, WANG Z P. Multi-time-scale framework for prognostic health condition of lithium battery using modified Gaussian process regression and nonlinear regression[J]. Journal of Power Sources, 2020, 467: doi: 10.1016/j.jpowsour.2020.228358.
22	FONTGALLAND G, PEDRO H J G. Normality and correlation coefficient in estimation of insulators' spectral signature[J]. IEEE Signal Processing Letters, 2015, 22(8): 1175-1179.

电池型号	恒流充电截止电压/V	恒压充电截止电流/A	恒流放电电流/A	恒流放电截止电压/V	电池容量/(A·h)	EOL/次
B0005	4.2	0.02	2	2.7	2	123
B0006	4.2	0.02	2	2.5	2	119
B0007	4.2	0.02	2	2.2	2	167
B0018	4.2	0.02	2	2.5	2	122

特征量	充电阶段		放电阶段				电池容量
特征量	D_C	T_C	V_D	D_D	T_D	T_S	Q
D_C	1	-0.37	-0.68	-0.69	0.75	-0.52	0.75
T_C	-0.37	1	0.59	0.66	-0.68	0.63	-0.68
V_D	-0.68	0.59	1	0.80	-0.88	0.69	-0.88
D_D	-0.69	0.66	0.80	1	-0.96	0.94	-0.96
T_D	0.75	0.68	-0.88	-0.96	1	-0.88	1
T_S	-0.52	0.63	0.69	0.94	-0.88	1	-0.89
Q	0.75	0.68	-0.88	-0.96	1	-0.89	1

方案号	间接健康因子	RMSE/%				MAE/%				时间/s
方案号	间接健康因子	B0005	B0006	B0007	B0018	B0005	B0006	B0007	B0018	B0005	B0006	B0007	B0018
1	D_C、V_D、T_D、T_S	2.930	1.838	1.723	1.344	2.441	1.615	1.311	1.126	395	422	409	389
2	T_D、T_S	1.825	2.824	2.473	2.014	1.483	2.126	1.902	1.742	992	1085	1337	1264
3	T_C、D_D、T_D、T_S	1.225	1.908	1.992	1.474	0.874	1.576	1.726	1.265	243	258	270	267
4	D_C、D_D	1.939	2.932	3.079	3.108	1.571	2.721	2.321	2.693	1211	1960	1826	1794
5	D_C、D_D、T_D、T_S	0.510	0.907	0.868	0.448	0.430	0.767	0.618	0.345	302	367	324	415

电池数据集	预测模型	RMSE/%	MAE/%	MSE/%
B0005	BP	2.729	2.508	0.007
B0005	Elman	0.410	0.350	0.002
B0006	BP	5.865	4.859	0.344
B0006	Elman	0.300	0.731	0.011
B0007	BP	3.303	3.007	0.109
B0007	Elman	0.575	0.469	0.003
B0018	BP	1.193	0.918	0.014
B0018	Elman	0.651	0.533	0.004

电池型号	预测起始点	预测周期数	MSE/%	RMSE/%	MAE/%	AE/周期数
B0005	90	77	0.195	4.420	3.360	3
	100	67	0.003	0.510	0.430	1
	110	57	0.003	0.550	0.450	1
	120	47	0.002	0.410	0.350	0
B0006	90	76	0.124	3.521	2.555	0
	100	66	0.955	3.090	2.450	0
	110	57	0.002	0.458	0.408	0
	120	46	0.001	0.300	0.731	—
B0007	90	77	0.038	1.957	1.668	0
	100	67	0.007	0.865	0.618	0
	110	57	0.002	0.424	0.347	0
	120	47	0.003	0.575	0.469	0
B0018	70	62	0.099	3.145	2.494	9
	80	52	0.058	2.399	2.029	9
	90	42	0.024	1.553	1.382	4
	100	32	0.004	0.651	0.533	0