基于优化Elman神经网络的锂电池容量预测

doi:10.19799/j.cnki.2095-4239.2021.0655

摘要/Abstract

摘要：

锂离子电池容量的精确预测有利于提升电池的使用安全和避免电池滥用，受其复杂的内部电化学反应和外部使用条件影响，对其老化后的容量精确预测一直是电池管理系统的难点之一。为实现锂电池全服役周期内容量的高效准确预测，本文提出了一种基于遗传算法优化的Elman神经网络(GA-Elman)电池容量预测模型。首先选择电池不同循环下的放电容量增量、内阻以及温度数据作为有效表征电池老化和容量衰减规律的特征量，其次运用主成分分析算法对特征量进行降维以降低训练量数据维度，然后基于Elman神经网络构建电池容量预测模型，并引入遗传算法优化Elman神经网络的权值和阈值，实现对电池容量的高效精确预测，最后在不同电池上对该模型进行了验证。验证结果表明：与传统Elman神经网络和长短期记忆神经网络(long and short term memory neural network，LSTM NN)预测模型相比，GA-Elman神经网络预测模型有更好的预测精度和更高的运算效率。在不同电池上该模型预测结果的最大平均绝对误差为0.92%，最大均方根误差为1.02%，最小拟合系数为0.9679，表明该模型可以精确预测锂电池衰退过程中的容量并且对不同电池有较强的适应性。

关键词: 老化特征, 遗传算法, 埃尔曼神经网络, 锂电池容量预测

Abstract:

Accurate prediction of lithium-ion battery capacity is advantageous for enhancing battery safety and avoiding battery abuse. An accurate capacity prediction has always been a challenge in battery management systems due to the effect of complicated internal electrochemical reactions and external variables. To achieve an effective and accurate prediction of lithium battery capacity in the full-service life, this study proposed an Elman neural network battery capacity prediction model optimized by a genetic algorithm. To begin, the typical characteristics of discharge capacity growth, internal resistance, and temperature data collected over various battery cycles were chosen to adequately depict the law of battery aging and capacity degradation. Secondly, the principal component analysis algorithm is used to reduce the dimension of the characteristic quantities to reduce the data dimension of the training quantity. The Elman neural network is then used to build the battery capacity prediction model. To ensure efficient and accurate prediction of battery capacity, a genetic algorithm is used to adjust the weights and thresholds of the Elman neural network. Finally, the model was verified on different batteries. The verification findings reveal that the GA-Elman neural network prediction model is more accurate and efficient than the classic Elman neural network and LSTM neural network prediction models. The maximum mean absolute error, maximum root mean square error, and minimum fitting coefficient of the model is 0.92%, 1.02%, and 0.9679, respectively, for various batteries, showing that the model can accurately predict the capacity of lithium battery in the process of decay and has high adaptability to different batteries.

Key words: aging characteristics, genetic algorithm, elman neural network, capacity prediction of lithium battery

中图分类号:

TM 911

黄鹏, 聂枝根, 陈峥, 舒星, 沈世全, 杨继鹏, 申江卫. 基于优化Elman神经网络的锂电池容量预测[J]. 储能科学与技术, 2022, 11(7): 2282-2294.

Peng HUANG, Zhigen NIE, Zheng CHEN, Xing SHU, Shiquan SHEN, Jipeng YANG, Jiangwei SHEN. Capacity prediction of lithium battery based on optimized Elman neural network[J]. Energy Storage Science and Technology, 2022, 11(7): 2282-2294.

图/表 17

表1

图1

图2

图3

图4

图5

图6

表2

图7

图8

图9

表3

图10

表4

图11

图12

表5

参考文献 26

1	CHEN Z, SUN M M, SHU X, et al. Online state of health estimation for lithium-ion batteries based on support vector machine[J]. Applied Sciences, 2018, 8(6): 925.
2	SHU X, LI G, SHEN J W, et al. An adaptive fusion estimation algorithm for state of charge of lithium-ion batteries considering wide operating temperature and degradation[J]. Journal of Power Sources, 2020, 462: doi: 10.1016/j.jpowsour.2020.228132.
3	HU X S, LI S E, YANG Y Lapproach. Advanced machine learning approach for lithium-ion battery state estimation in electric vehicles[J]. IEEE Transactions on Transportation Electrification, 2016, 2(2): 140-149 .
4	马彦,陈阳,张帆,等. 基于扩展H∞粒子滤波算法的动力电池寿命预测方法[J]. 机械工程学报, 2019, 55(20): 36-43.
	MA Y, CHEN Y, ZHANG F, et al. Remaining useful life prediction of power battery based on extend H∞ particle filter algorithm[J]. Journal of Mechanical Engineering, 2019, 55(20): 36-43.
5	HU X S, YUAN H, ZOU C F, et al. Co-estimation of state of charge and state of health for lithium-ion batteries based on fractional-order calculus[J]. IEEE Transactions on Vehicular Technology, 2018, 67(11): 10319-10329.
6	SAHA B, GOEBEL K, POLL S, et al. Prognostics methods for battery health monitoring using a Bayesian framework[J]. IEEE Transactions on Instrumentation and Measurement, 2009, 58(2): 291-296.
7	韦海燕, 安晶晶, 陈静, 等. 基于改进粒子滤波算法实现锂离子电池RUL预测[J]. 汽车工程, 2019, 41(12): 1377-1383.
	WEI H Y, AN J J, CHEN J, et al. RUL prediction of lithium-ion battery based on improved particle filtering algorithm[J]. Automotive Engineering, 2019, 41(12): 1377-1383.
8	郑涛, 张里, 侯杨成, 等. 基于自适应CKF的老化锂电池SOC估计[J]. 储能科学与技术, 2020, 9(4): 1193-1199.
	ZHENG T, ZHANG L, HOU Y C, et al. SOC estimation of aging lithium battery based on adaptive CKF[J]. Energy Storage Science and Technology, 2020, 9(4): 1193-1199.
9	NG S S Y, XING Y J, TSUI K L. A naive Bayes model for robust remaining useful life prediction of lithium-ion battery[J]. Applied Energy, 2014, 118: 114-123.
10	ZHANG Y Z, XIONG R, HE H W, et al. Lithium-ion battery remaining useful life prediction with box-cox transformation and Monte Carlo simulation[J]. IEEE Transactions on Industrial Electronics, 2019, 66(2): 1585-1597.
11	DAI H D, ZHAO G C, LIN M Q, et al. A novel estimation method for the state of health of lithium-ion battery using prior knowledge-based neural network and Markov chain[J]. IEEE Transactions on Industrial Electronics, 2019, 66(10): 7706-7716.
12	DENG Y W, YING H J, ER J Q, et al. Feature parameter extraction and intelligent estimation of the state-of-health of lithium-ion batteries[J]. Energy, 2019, 176: 91-102.
13	韩云飞, 谢佳, 蔡涛, 等. 结合高斯过程回归与特征选择的锂离子电池容量估计方法[J]. 储能科学与技术, 2021, 10(4): 1432-1438.
	HAN Y F, XIE J, CAI T, et al. Capacity estimation of lithium-ion batteries based on Gaussian process regression and feature selection[J]. Energy Storage Science and Technology, 2021, 10(4): 1432-1438.
14	张新锋, 姚蒙蒙, 王钟毅, 等. 基于ACO-BP神经网络的锂离子电池容量衰退预测[J]. 储能科学与技术, 2020, 9(1): 138-144.
	ZHANG X F, YAO M M, WANG Z Y, et al. Lithium-ion battery capacity decline prediction based on ant colony optimization BP neural network algorithm[J]. Energy Storage Science and Technology, 2020, 9(1): 138-144.
	张新锋, 姚蒙蒙, 王钟毅, 等. 基于ACO-BP神经网络的锂离子电池容量衰退预测[J]. 储能科学与技术, 2020(1): 173-179.
	ZHANG X F, YAO M M, WANG Z Y, et al. Lithium-ion battery capacity decline prediction based on ant colony optimization BP neural network algorithm[J]. Energy Storage Science and Technology, 2020(1): 173-179.
15	GUO P Y, CHENG Z, YANG L. A data-driven remaining capacity estimation approach for lithium-ion batteries based on charging health feature extraction[J]. Journal of Power Sources, 2019, 412: 442-450.
16	ZHOU D, YIN H T, XIE W, et al. Research on online capacity estimation of power battery based on EKF-GPR model[J]. Journal of Chemistry, 2019: doi:10.1155/2019/5327319.
17	LI X Y, WANG Z P, YAN J Y. Prognostic health condition for lithium battery using the partial incremental capacity and Gaussian process regression[J]. Journal of Power Sources, 2019, 421: 56-67.
18	LI X Y, ZHANG L, WANG Z P, et al. Remaining useful life prediction for lithium-ion batteries based on a hybrid model combining the long short-term memory and Elman neural networks[J]. Journal of Energy Storage, 2019, 21: 510-518.
19	SEVERSON K A, ATTIA P M, JIN N, et al. Data-driven prediction of battery cycle life before capacity degradation[J]. Nature Energy, 2019, 4(5): 383-391.
20	LI Y, ABDEL-MONEM M, GOPALAKRISHNAN R, et al. A quick on-line state of health estimation method for Li-ion battery with incremental capacity curves processed by Gaussian filter[J]. Journal of Power Sources, 2018, 373: 40-53.
21	LI X Y, WANG Z P, ZHANG L, et al. State-of-health estimation for Li-ion batteries by combing the incremental capacity analysis method with grey relational analysis[J]. Journal of Power Sources, 2019, 410/411: 106-114.
22	WANG P, MENG P, SONG B W. Response surface method using grey relational analysis for decision making in weapon system selection[J]. Journal of Systems Engineering and Electronics, 2014, 25(2): 265-272.
23	LU C, TAO L F, FAN H Z. Li-ion battery capacity estimation: A geometrical approach[J]. Journal of Power Sources, 2014, 261: 141-147.
24	SUN W Z, WANG J S. Elman neural network soft-sensor model of conversion velocity in polymerization process optimized by chaos whale optimization algorithm[J]. IEEE Access, 2017, 5: 13062-13076.
25	LIU H, TIAN H Q, LIANG X F, et al. Wind speed forecasting approach using secondary decomposition algorithm and Elman neural networks[J]. Applied Energy, 2015, 157: 183-194.
26	HE W, WILLIARD N, OSTERMAN M, et al. Prognostics of lithium-ion batteries based on dempster-shafer theory and the bayesian monte carlo method[J]. Journal of Power Sources, 2011, 196(23): 10314-10321.

电池编号	特征量
电池编号	F1	F2	F3	F4
电池1	0.931	0.885	0.693	0.832
电池2	0.948	0.893	0.764	0.811
电池3	0.936	0.872	0.782	0.829
电池4	0.927	0.884	0.685	0.823

电池误差评估	电池1训练量			电池2训练量			电池3训练量			电池4训练量
电池误差评估	40%	60%	80%	40%	60%	80%	40%	60%	80%	40%	60%	80%
R²	0.933	0.988	0.995	0.953	0.983	0.991	0.962	0.971	0.984	0.945	0.974	0.987
MAE/%	1.2	0.21	0.13	0.84	0.24	0.15	0.72	0.35	0.14	1.03	0.29	0.13
RMSE/%	1.97	0.41	0.24	1.75	0.54	0.36	1.53	0.93	0.57	1.86	0.75	0.44

不同方法	电池1		电池2		电池3		电池4
不同方法	RMSE/%	时间/s	RMSE/%	时间/s	RMSE/%	时间/s	RMSE/%	时间/s
LSTM NN	0.89/0.76	20.95/21.57	0.76/0.99	21.14/21.77	1.09/1.14	22.53/23.25	0.82/0.87	21.76/22.41
Elman NN	1.51/1.19	13.15/13.22	1.29/1.63	13.62/13.74	1.49/1.35	13.79/13.92	1.22/1.43	13.12/13.25
GA-Elman NN	0.66/0.41	10.79/10.73	0.57/0.33	10.51/10.39	0.34/0.38	11.16/11.02	0.43/0.46	10.14/10.25

电池	R²	MAE/%	RMSE/%
2	0.9679	0.92	1.02
3	0.9717	0.78	0.89
4	0.9863	0.53	0.61
5	0.9615	1.03	1.26
6	0.9587	1.15	1.47

[1]	韩帅, 孙乐平, 卢健斌, 郭小璇. 含电动汽车的气电互联虚拟电厂区间多目标优化调度策略[J]. 储能科学与技术, 2022, 11(5): 1428-1436.
[2]	张德隆, MUBAARAK Saif, 蒋思宇, 王龙泽, 刘金鑫, 陈永聪, 李美成. 基于概率潮流的光伏电站中储能系统的优化配置方法[J]. 储能科学与技术, 2021, 10(6): 2244-2251.
[3]	虞启辉, 田利, 李晓飞, 李晓东, 谭心, 张业明. 考虑风能不确定性的压缩空气储能容量配置及经济性评估[J]. 储能科学与技术, 2021, 10(5): 1614-1623.
[4]	胡银全, 刘和平. 基于遗传算法的电池包高效热管理流道优化[J]. 储能科学与技术, 2021, 10(4): 1446-1453.
[5]	陆燕娟, 陈友芹, 潘庭龙. 考虑EV和需求侧响应的社区微网能量管理[J]. 储能科学与技术, 2021, 10(2): 617-623.
[6]	王泰华, 张书杰, 陈金干. 锂离子电池低温充电老化建模及其充电策略优化[J]. 储能科学与技术, 2020, 9(4): 1137-1146.
[7]	夏新茂, 关洪浩, 丁鹏飞, 孟高军. 基于改进型量子遗传算法的储能系统容量配置与优化策略[J]. 储能科学与技术, 2019, 8(3): 551-558.