多尺度分解下GRU-TCN集成的动力电池剩余使用寿命预测方法

doi:10.19799/j.cnki.2095-4239.2023.0754

储能科学与技术 ›› 2024, Vol. 13 ›› Issue (3): 1009-1018.doi: 10.19799/j.cnki.2095-4239.2023.0754

多尺度分解下GRU-TCN集成的动力电池剩余使用寿命预测方法

刘佳¹(), 马志强¹^,³(), 刘广忱²^,³, 高俊东¹, 李宏勋¹

^1.内蒙古工业大学数据科学与应用学院
^2.内蒙古工业大学大学电力学院
^3.大规模储能技术;教育部工程研究中心，内蒙古呼和浩特 010080

收稿日期:2023-10-30 修回日期:2023-11-08 出版日期:2024-03-28 发布日期:2024-03-28
通讯作者: 马志强 E-mail:liu_jia0429@126.com;mzq_bim@imut.edu.cn
作者简介:刘佳（2000—），女，硕士研究生，研究方向为深度学习、时间序列预测，E-mail：liu_jia0429@126.com；
基金资助:
国家自然基金(62166029);内蒙古自治区高等学校碳达峰碳中和研究项目(STZX202307)

Predicting the residual useful life of power batteries based on the GRUU-TCN ensemble under multiscale decomposition

Jia LIU¹(), Zhiqiang MA¹^,³(), Guangchen LIU²^,³, Jundong GAO¹, Hongxun LI¹

^1.College of Data Science and Application, Inner Mongolia University of Technology
^2.College of Electric Power, Inner Mongolia University of Technology
^3.Engineering Research Center of large-scale Energy Storage Technology, Ministry of Education, Hohhot 010080, Inner Mongolia, China

Received:2023-10-30 Revised:2023-11-08 Online:2024-03-28 Published:2024-03-28
Contact: Zhiqiang MA E-mail:liu_jia0429@126.com;mzq_bim@imut.edu.cn

摘要/Abstract

摘要：

精准预测动力电池的剩余使用寿命（remaining useful life，RUL）能够提前规避因电池过度使用带来的风险，为退役电池的二次利用提供决策依据，提升电池第二寿命的利用率。为了降低动力电池RUL预测任务中噪声和容量回升现象导致的非线性特征对RUL预测精度的影响，提出了一种基于集合经验模态分解（ensemble empirical mode decomposition，EEMD）、门控循环单元网络（gated recurrent unit，GRU）和时序卷积网络（temporal convolutional networks，TCN）集成的动力电池RUL预测模型。首先，使用EEMD对原始数据进行分解，动力电池容量衰退过程中由噪声和容量回升现象导致的非线性特征被分解到高频分量，而原始容量数据的主要趋势被分解到低频分量。其次，再使用GRU和TCN网络分别对高频分量和低频分量进行预测。最后，使用Attention对预测结果进行集成。在NASA数据集上的实验结果表明，本工作提出的集成模型的预测精度和对非线性特征的拟合程度都优于其他单一模型和其他同类型模型，最大平均绝对误差和最大均方根误差分别在0.52%和0.74%内，绝对误差在1个循环周期内，证明本模型有较好的RUL预测能力。

关键词: 动力电池, 剩余使用寿命, 经验模态分解, 门控循环单元网络, 时序卷积网络

Abstract:

Accurate prediction of the remaining useful life (RUL) of power batteries can avoid the risk of battery overuse, inform decision making on the secondary use of retired batteries, and improve the utilization rate of second-life batteries. We propose a method based on ensemble empirical mode decomposition (EEMD), gated recurrent units (GRUs), and temporal recurrent unit networks to reduce the dependence of RUL prediction accuracy on nonlinear features (this dependence is caused by noise and capacity recovery in the power battery RUL prediction task). GRU and temporal convolutional networks (TCNs) are integrated into the RUL prediction model for power batteries. First, the raw data are decomposed using EEMD, and the nonlinear features caused by noise and capacity rebound during power battery capacity decline are decomposed into high-frequency components. The main trends of the raw capacity data are decomposed into low-frequency components. Next, GRUs and TCNs are used to predict the high- and low-frequency components, respectively. Finally, the predictions are integrated using attention. The experimental results on the NASA dataset show that the prediction accuracy and the fitting of nonlinear features of the integrated model proposed in this paper are better than those of other single models and other models of the same type, with the maximum average absolute error and the maximum root-mean-square error within 0.52% and 0.74%, respectively, and the absolute error within one cycle period. These results prove that the proposed model produced more accurate predictions of the RUL than conventional models.

Key words: power battery, remaining service life, empirical mode decomposition, gated recurrent unit network, temporal convolutional networks

中图分类号:

TM 912

刘佳, 马志强, 刘广忱, 高俊东, 李宏勋. 多尺度分解下GRU-TCN集成的动力电池剩余使用寿命预测方法[J]. 储能科学与技术, 2024, 13(3): 1009-1018.

Jia LIU, Zhiqiang MA, Guangchen LIU, Jundong GAO, Hongxun LI. Predicting the residual useful life of power batteries based on the GRUU-TCN ensemble under multiscale decomposition[J]. Energy Storage Science and Technology, 2024, 13(3): 1009-1018.

图/表 7

图1

图2

图3

图4

表1

表2

表3

参考文献 20

1	武明虎, 岳程鹏, 张凡, 等. 多尺度分解下GRU-MLR组合的锂电池剩余使用寿命预测方法[J]. 储能科学与技术, 2023, 12(7): 2220-2228.
	WU M H, YUE C P, ZHANG F, et al. Combined GRU-MLR method for predicting the remaining useful life of lithium batteries via multiscale decomposition[J]. Energy Storage Science and Technology, 2023, 12(7): 2220-2228.
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: 228861.
3	WANG D, KONG J Z, YANG F F, et al. Battery prognostics at different operating conditions[J]. Measurement, 2020, 151: 107182.
4	GUHA A, PATRA A, VAISAKH K V. Remaining useful life estimation of lithium-ion batteries based on the internal resistance growth model[C]//2017 Indian Control Conference (ICC). Guwahati, India. IEEE, 2017: 33-38.
5	ZHANG H, MIAO Q, ZHANG X, et al. An improved unscented particle filter approach for lithium-ion battery remaining useful life prediction[J]. Microelectronics Reliability, 2018, 81: 288-298.
6	王萍, 范凌峰, 程泽. 基于健康特征参数的锂离子电池SOH和RUL联合估计方法[J]. 中国电机工程学报, 2022, 42(4): 1523-1533, 25.
	WANG P, FAN L F, CHENG Z. A joint state of health and remaining useful life estimation approach for lithium-ion batteries based on health factor parameter[J]. Proceedings of the CSEE, 2022, 42(4): 1523-1533, 25.
7	廖力, 肖廷奕, 吴铁洲, 等. 基于多健康特征融合的锂电池SOH和RUL预测[J]. 电源技术, 2023, 47(2): 193-198.
	LIAO L, XIAO T Y, WU T Z, et al. SOH and RUL prediction for lithium batteries based on fusion of multiple health features[J]. Chinese Journal of Power Sources, 2023, 47(2): 193-198.
8	吴忠强, 胡晓宇, 马博岩, 等. 基于 PF-LSTM 的锂电池剩余使用寿命预测[J]. 计量学报, 2023, 44(6): 939-947.
	WU Z Q, HU X Y, MA B Y, et al. Residual service life prediction of lithium batteries based on PF-LSTM [J]. Acta Metrologica Sinica, 2023, 44(6): 939-947.
9	CHEN D Q, HONG W C, ZHOU X Z. Transformer network for remaining useful life prediction of lithium-ion batteries[J]. IEEE Access, 1975, 10: 19621-19628.
10	HONG J, LEE D, JEONG E R, et al. Towards the swift prediction of the remaining useful life of lithium-ion batteries with end-to-end deep learning[J]. Applied Energy, 2020, 278: 115646.
11	董渊昌, 庞晓琼, 贾建芳, 等. 基于SVD-SAE-GPR的锂离子电池RUL预测[J]. 储能科学与技术, 2023, 12(4): 1257-1267.
	DONG Y C, PANG X Q, JIA J F, et al. Remaining useful life prediction of lithium-ion batteries based on SVD-SAE-GPR[J]. Energy Storage Science and Technology, 2023, 12(4): 1257-1267.
12	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, 2021, 68(4): 3170-3180.
13	MAO L, XU J, CHEN J J, et al. A LSTM-STW and GS-LM fusion method for lithium-ion battery RUL prediction based on EEMD[J]. Energies, 2020, 13(9): 2380.
14	WU Z H, HUANG N E. Ensemble empirical mode decomposition: A noise-assisted data analysis method[J]. Advances in Adaptive Data Analysis, 2009, 1(1): 1-41.
15	BAI S J, KOLTER J Z, KOLTUN V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling[J]. arXiv preprint arXiv:, 2018.
16	CHO K, MERRIENBOER B V, GÜLÇEHRE Ç, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[J]. arXiv preprint arXiv:, 2014.
17	FUKUSHIMA K. Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position[J]. Biological Cybernetics, 1980, 36(4): 193-202.
18	RUMELHART D E, HINTON G E, WILLIAMS R J. Learning representations by back-propagating errors[J]. Nature, 1986, 323: 533-536.
19	BAHDANAU D, CHO K, BENGIO Y. Neural machine translation by jointly learning to align and translate[J]. arXiv preprint arXiv:, 2014.
20	SAHA B, GOEBEL K. Battery data set[R]. NASA Ames Prognostics Data Repository, 2007.

电池	预测起点	方法	RUL预测值	AE	MAE	RMSE
B0005	80	GRU	35	10	0.0110	0.0135
		LSTM	36	9	0.0096	0.0124
		TCN	42	3	0. 0082	0.0142
		EEMD-GRU-ARIMA	35	10	0.0068	0.0092
		EEMD-GRU-MLR	34	11	0.0067	0.0073
		EEMD-GRU-GRU	35	10	0.0075	0.0082
		EEMD-GRU-TCN	36	9	0.0054	0.0070
		EEMD-GRU-TCN-Stacking	38	7	0.0057	0.0077
		Ours	44	1	0.0052	0.0074

B0006	80	GRU	29	1	0.0116	0.0127
		LSTM	30	0	0.0111	0.0114
		TCN	29	1	0.0128	0.0222
		EEMD-GRU-ARIMA	27	3	0.0067	0.0091
		EEMD-GRU-MLR	17	13	0.0079	0.0149
		EEMD-GRU-GRU	26	4	0.0086	0.0116
		EEMD-GRU-TCN	30	0	0.0053	0.0066
		EEMD-GRU-TCN-Stacking	16	14	0.0113	0.0178
		Ours	30	0	0.0047	0.0050

B0007	80	GRU	68	12	0.0112	0.0117
		LSTM	69	11	0.0104	0.0111
		TCN	82	2	0.0105	0.0164
		EEMD-GRU-ARIMA	80	0	0.0062	0.0103
		EEMD-GRU-MLR	76	4	0.0068	0.0074
		EEMD-GRU-GRU	75	5	0.0099	0.0120
		EEMD-GRU-TCN	82	2	0.0052	0.0072
		EEMD-GRU-TCN-Stacking	65	15	0.0058	0.0076
		Ours	80	0	0.0047	0.0072

电池	预测起点	RUL预测值	AE	MAE	RMSE
B0005	70	51	4	0.0103	0.0110
	80	44	1	0.0052	0.0074
	90	35	0	0.0033	0.0044
	100	25	0	0.0030	0.0043

B0006	70	37	3	0.0104	0.0138
	80	30	0	0.0047	0.0050
	90	20	0	0.0043	0.0054
	100	10	0	0.0039	0.0052

B0007	70	88	2	0.0112	0.0145
	80	80	0	0.0047	0.0072
	90	70	0	0.0030	0.0036
	100	60	0	0.0018	0.0026

电池	预测起点	预测方法			RUL预测值	AE	MAE	RMSE
电池	预测起点	GRU	TCN	Attention	RUL预测值	AE	MAE	RMSE
B0005	80	√			—	—	0.7064	0.7076
			√		44	1	0.0061	0.0080
		√	√		44	1	0.0054	0.0077
		√		√	—	—	0.7055	0.7067
			√	√	42	3	0.0067	0.0081
		√	√	√	44	1	0.0052	0.0074

B0006	80	√			—	—	0.6743	0.6760
			√		31	1	0.0088	0.0127
		√	√		28	2	0.0053	0.0066
		√		√	—	—	0.6730	0.6746
			√	√	38	8	0.0114	0.0148
		√	√	√	30	0	0.0047	0.0050

B0007	80	√			—	—	0.7542	0.7549
			√		81	1	0.0058	0.0077
		√	√		80	0	0.0052	0.0072
		√		√	—	—	0.7552	0.7559
			√	√	82	2	0.0108	0.0120
		√	√	√	80	0	0.0047	0.0072

[1]	梁宏毅, 陈锋, 甘友毅, 邵丹. 动力锂电池三元正极低温性能研究[J]. 储能科学与技术, 2024, 13(1): 293-298.
[2]	武明虎, 岳程鹏, 张凡, 李俊晓, 黄伟, 胡胜, 唐靓. 多尺度分解下GRU-MLR组合的锂电池剩余使用寿命预测方法[J]. 储能科学与技术, 2023, 12(7): 2220-2228.
[3]	赵光金, 李博文, 胡玉霞, 董锐锋, 王放放. 退役动力电池梯次利用技术及工程应用概述[J]. 储能科学与技术, 2023, 12(7): 2319-2332.
[4]	于会群, 胡哲豪, 彭道刚, 孙浩益. 退役动力电池回收及其在储能系统中梯次利用关键技术[J]. 储能科学与技术, 2023, 12(5): 1675-1685.
[5]	董渊昌, 庞晓琼, 贾建芳, 史元浩, 温杰, 李笑, 张鑫. 基于SVD-SAE-GPR的锂离子电池RUL预测[J]. 储能科学与技术, 2023, 12(4): 1257-1267.
[6]	刘峰, 陈海忠. 基于CEEMDAN和ISOA-ELM的锂电池荷电状态预测[J]. 储能科学与技术, 2023, 12(4): 1244-1256.
[7]	王朋凯, 张新燕, 张光昊. 基于ResNet-Bi-LSTM-Attention的锂离子电池剩余使用寿命预测[J]. 储能科学与技术, 2023, 12(4): 1215-1222.
[8]	厉运杰, 张光雨, 祝维文, 闵远远, 饶成飞, 孙言飞, 徐庆庆. 基于壅塞流的动力电池防爆阀泄压特性的动态仿真[J]. 储能科学与技术, 2023, 12(3): 960-967.
[9]	魏震波, 姚怡欣, 张雯雯, 罗紫航, 李银江, 任语杰. 基于完备集合经验模态分解的含抽蓄微电网混合储能容量优化配置[J]. 储能科学与技术, 2023, 12(11): 3414-3424.
[10]	陈锐, 丁凯, 祖连兴, 许青松, 王宗标, 罗大思, 苏敬江, 胡圣, 毛冀龙. 基于AED-CEEMD-Transformer的锂离子电池健康状态估计[J]. 储能科学与技术, 2023, 12(10): 3242-3253.
[11]	李练兵, 朱乐, 景睿雄, 王兰超, 韩琪琪. 基于DESSA-DESN和NCA的锂离子电池剩余寿命预测[J]. 储能科学与技术, 2023, 12(10): 3191-3202.
[12]	宋兴海, 张小乾, 梁惠施, 史梓男, 李棉刚, 周奎, 贡晓旭. 基于SDAE-Transformer-ECA网络的锂电池剩余使用寿命预测[J]. 储能科学与技术, 2023, 12(10): 3181-3190.
[13]	常泽宇, 张之琦, 张晓东, 李丽, 郁亚娟. 基于数据驱动的动力电池健康状态评估平台[J]. 储能科学与技术, 2022, 11(6): 1847-1853.
[14]	周伟, 符冬菊, 刘伟峰, 陈建军, 胡照, 曾燮榕. 废旧磷酸铁锂动力电池回收利用研究进展[J]. 储能科学与技术, 2022, 11(6): 1854-1864.
[15]	王苏杭, 李建林, 李雅欣, 熊俊杰, 曾伟. 锂离子电池系统低温充电策略[J]. 储能科学与技术, 2022, 11(5): 1537-1542.

多尺度分解下GRU-TCN集成的动力电池剩余使用寿命预测方法

Predicting the residual useful life of power batteries based on the GRUU-TCN ensemble under multiscale decomposition

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 20

相关文章 15

编辑推荐

Metrics

本文评价