基于GWO-LSTM与LSSVM的锂离子电池荷电状态与容量联合估计

doi:10.3969/j.issn.1674-8484.2022.03.019

汽车安全与节能学报 ›› 2022, Vol. 13 ›› Issue (3): 571-579.DOI: 10.3969/j.issn.1674-8484.2022.03.019

基于GWO-LSTM与LSSVM的锂离子电池荷电状态与容量联合估计

王桥¹(), 魏孟¹^,²(), 叶敏¹^,^*(), 廉高棨¹, 麻玉川¹

1.长安大学公路养护装备国家工程实验室,西安 710064,中国
2.新加坡国立大学机械工程系,117576, 新加坡

收稿日期:2021-11-25 修回日期:2022-05-11 出版日期:2022-09-30 发布日期:2022-10-04
通讯作者: 叶敏
作者简介:* 叶敏(1978—),男(汉),吉林,教授。E-mail： mingye@chd.edu.cn。
王桥(1996—),男(汉),陕西,博士研究生。E-mail： qiaowang@chd.edu.cn。
魏孟(1997—),男(汉),陕西,博士研究生。E-mail： 2019025006@chd.edu.cn。
基金资助:
中央高校基本科研业务费专项资金-长安大学优秀博士学位论文培育资助项目(300102252710);长安大学研究生科研创新实践项目(300103722004)

Co-estimation of state of charge and capacity of lithium-ion battery based on GWO optimized LSTM and LSSVM

WANG Qiao¹(), WEI Meng¹^,²(), YE Min¹^,^*(), LIAN Gaoqi¹, MA Yuchuan¹

1. National Engineering Laboratory for Highway Maintenance Equipment of Chang ‘an University, Xi’an 710064, China
2. Department of Mechanical Engineering of National University of Singapore, 117576, Singapore

Received:2021-11-25 Revised:2022-05-11 Online:2022-09-30 Published:2022-10-04
Contact: YE Min

摘要/Abstract

摘要：

为了提高锂离子电池老化后的荷电状态(SOC)估计精度,通过分析锂离子电池的充电与放电特性,提出一种基于长短时记忆(LSTM)网络和最小二乘支持向量机(LSSVM)的荷电状态与容量联合估计模型。根据锂离子电池的充放电特性,提出片段电压的充电时间作为健康因子;基于最小二乘支持向量机建立了锂离子电池的容量估计模块,容量估计结果通过记忆门控被记录下来;基于灰狼算法优化的长短时记忆网络(GWO-LSTM)框架建立了锂离子电池的荷电状态与容量的联合估计模型。结果表明：与粒子群算法优化的反向传播神经网络(BPNN-PSO)和传统长短时记忆网络模型对比,所提方法的容量估计精度提高了43%以上,SOC估计表现出更好的鲁棒性。

关键词: 锂离子电池, 荷电状态(SOC)估计, 容量估计, 长短时记忆网络(LSTM), 灰狼优化(GWO), 最小二乘支持向量机(LSSVM)

Abstract:

A joint estimation model of the state of charge and capacity based on long short term memory (LSTM) and least squares support vector machine (LSSVM) was proposed to improve the accuracy of the state of charge (SOC) estimation of lithium-ion batteries after aging by analyzing the charging and discharging characteristics of lithium-ion batteries. According to the charging and discharging characteristics of the lithium-ion batteries, the charging time of the segment voltage was proposed as a new health factor. The capacity estimation module of the lithium-ion battery was established based on LSSVM, and the capacity estimation result was recorded by a memory gate. A joint estimation model of SOC and capacity of lithium-ion battery was established based on the grey wolf optimizer optimized long short term memory (GWO-LSTM) framework. The results show that the capacity estimation accuracy of the proposed method is improved by more than 43% comparing with the back-propagation neural network optimized by the particle swarm algorithm (BPNN-PSO) and the traditional LSTM model, and the SOC estimation shows better robustness.

Key words: lithium-ion batteries, state of charge (SOC), capacity estimation, long short memory (LSTM), grey wolf optimizer(GWO), least square support vector machine (LSSVM)

中图分类号:

TM911.3

王桥, 魏孟, 叶敏, 廉高棨, 麻玉川. 基于GWO-LSTM与LSSVM的锂离子电池荷电状态与容量联合估计[J]. 汽车安全与节能学报, 2022, 13(3): 571-579.

WANG Qiao, WEI Meng, YE Min, LIAN Gaoqi, MA Yuchuan. Co-estimation of state of charge and capacity of lithium-ion battery based on GWO optimized LSTM and LSSVM[J]. Journal of Automotive Safety and Energy, 2022, 13(3): 571-579.

图/表 8

图1 测试样本实验数据

测试样本	Spearman相关性系数
测试样本	3.6~3.7 V	3.7~3.8 V	3.8~3.9 V	3.9~4.0 V	完整充电时间
CX2-1	0.699 3	0.662 1	0.912 6	0.963 5	0.957 6
CX2-2	0.700 2	0.651 2	0.927 1	0.978 5	0.972 5
CX2-3	0.733 2	0.654 3	0.931 1	0.981 4	0.977 5

表1 Spearman相关性分析结果

图2 LSTM结构图

图3 联合估计流程图

图4 实验测试平台

图5 CX2-3与CX2-3容量估计结果

图6 CX2-2 的SOC估计结果曲线

图7 CX2-3 的SOC估计结果

参考文献 22

[1]	CHEN Cheng, XIONG Rui, SHEN Weixiang. A lithium-ion battery-in-the-loop approach to test and validate multiscale dual H infinity filters for state-of-charge and capacity estimation[J]. IEEE Trans Power Electr, 2017, 33: 332-342. doi: 10.1109/TPEL.2017.2670081 URL
[2]	Shrivastava P, Soon T K, Idris M, et al. Overview of model-based online state-of-charge estimation using Kalman filter family for lithium-ion batteries[J]. Rene Sustain Energ Rev, 2019, 113: 109233.
[3]	WANG Yujie, TIAN Jiaqiang, SUN Zhendong, et al. A comprehensive review of battery modeling and state estimation approaches for advanced battery management systems[J]. Rene Sustain Energ Rev, 2020, 131: 110015.
[4]	XIONG Rui, CAO Jiayi, YU Quanqing, et al. Critical review on the battery state of charge estimation methods for electric vehicles[J]. IEEE Access, 2017, 6: 1832-1843. doi: 10.1109/ACCESS.2017.2780258 URL
[5]	XIONG Rui, YU Quanqing, WANG Leyi, et al. A novel method to obtain the open circuit voltage for the state of charge of lithium ion batteries in electric vehicles by using H infinity filter[J]. Appl Energ, 2017, 207: 346-353. doi: 10.1016/j.apenergy.2017.05.136 URL
[6]	张宏伟, 李建成, 张兵兵. 基于安时积分和粒子滤波修正的锂电池 SOC 估计[J]. 传感器与微系统, 2016, 35: 4-7.
	ZHANG Hongwei, LI Jiancheng, ZHANG Bingbing. SOC estimation of lithium battery based on Ampere-hour integral and particle filter[J]. Sens Microsyst, 2016, 35: 4-7. (in Chinese)
[7]	王晓兰, 靳皓晴, 刘祥远. 基于融合模型的锂离子电池荷电状态在线估计[J]. 工程科学学报, 2020, 42: 1200-1208.
	WANG Xiaolan, JIN Haoqing, LIU Xiangyuan. Online estimation of lithium-ion battery charge state based on fusion model[J]. Chin J Engi, 2020(42): 1200-1208. (in Chinese)
[8]	张远进, 吴华伟, 叶从进. 基于 AUKF-BP 神经网络的锂电池 SOC 估算[J]. 储能科学与技术, 2021, 10: 237-241.
	ZHANG Yuanjin, WU Huawei, YE Congjin. SOC estimation of lithium battery based on AUKF-BP neural network[J]. Energ Stor Sci Tech, 2021, 10: 237-241. (in Chinese)
[9]	CHE Yanbo, LIU Yushu, CHENG Ze, et al. SOC and SOH identification method of Li-ion battery based on SWPSO-DRNN[J]. IEEE J Emerg Sele Top Power Electr, 2020, 9(4): 4050-4061.
[10]	YANG Fangfang, ZHANG Shaohui, LI Weihua, et al. State-of-charge estimation of lithium-ion batteries using LSTM and UKF[J]. Energy, 2020, 201: 117664. doi: 10.1016/j.energy.2020.117664 URL
[11]	LI Junhui, GAO Fengjie, YAN Gangui, et al. Modeling and SOC estimation of lithium iron phosphate battery considering capacity loss[J]. Prot Contr Mode Power Syst, 2018, 3: E1-9.
[12]	卢婷, 杨文强. 锂离子电池全生命周期内评估参数及评估方法综述[J]. 储能科学与技术, 2020, 9: 657-669.
	LU Ting, YANG Wenqiang. Review of evaluation parameters and methods of lithium batteries throughout its life cycle[J]. Energ Stor Sci Tech, 2020, 9: 657-669. (in Chinese)
[13]	Park J, Kim K, Park S, et al. Complementary cooperative SOC/Capacity estimator based on the discrete variational derivative combined with the DEKF for electric power applications[J]. Energy, 2021, 232: 121023. doi: 10.1016/j.energy.2021.121023 URL
[14]	章军辉, 李庆, 陈大鹏, 等. 基于快速 S, R-UKF 的锂离子动力电池 SOC 联合估计[J]. 工程科学学报, 2021, 43: 1-9.
	ZHANG Junhui, LI Qing, CHEN Dapeng, et al. Real-time SOC co-estimation algorithm for li-ion batteries based on fast square root unscented kalman filter[J]. Chin J Engi, 2021, 43: 1-9. (in Chinese)
[15]	HAN Zhezhe, Hossain M M, WANG Yuwei, et al. Combustion stability monitoring through flame imaging and stacked sparse autoencoder based deep neural network[J]. Appl Energ, 2020, 259: 114159. doi: 10.1016/j.apenergy.2019.114159 URL
[16]	LAI Xin, YI Wei, CUI Yifan, et al. Capacity estimation of lithium-ion cells by combining model-based and data-driven methods based on a sequential extended Kalman filter[J]. Energy, 2021, 216: 119233. doi: 10.1016/j.energy.2020.119233 URL
[17]	史永胜, 施梦琢, 丁恩松, 等. 基于 CEEMDAN-LSTM 组合的锂离子电池寿命预测方法[J]. 工程科学学报, 2021, 43(7): 1-10.
	SHI Yongsheng, SHI Mengzhuo, DING Ensong, et al. Online estimation of the state of charge of a lithium-ion battery based on the fusion model[J]. Chin J Engi, 2021, 43(7): 985-994. (in Chinese)
[18]	LI Xiaoyu, YUAN Changgui, WANG Zhenpo. State of health estimation for Li-ion battery via partial incremental capacity analysis based on support vector regression[J]. Energy, 2020, 203: 117852. doi: 10.1016/j.energy.2020.117852 URL
[19]	DENG Zhongwei, YANG Lin, CAI Yishan, et al. Online available capacity prediction and state of charge estimation based on advanced data-driven algorithms for lithium iron phosphate battery[J]. Energy, 2016, 112: 469-480. doi: 10.1016/j.energy.2016.06.130 URL
[20]	JIANG Bo, DAI Haifeng, WEI Xuezhe. Incremental capacity analysis based adaptive capacity estimation for lithium-ion battery considering charging condition[J]. Appl Energ, 2020, 269: 115074. doi: 10.1016/j.apenergy.2020.115074 URL
[21]	QIAN Cheng, XU Binghui, CHANG Liang, et al. Convolutional neural network based capacity estimation using random segments of the charging curves for lithium-ion batteries[J]. Energy, 2021, 227: 120333. doi: 10.1016/j.energy.2021.120333 URL
[22]	PAN Hailong, LU Zhiqiang, WANG Huimin, et al. Novel battery state-of-health online estimation method using multiple health indicators and an extreme learning machine[J]. Energy, 2018, 160: 466-77. doi: 10.1016/j.energy.2018.06.220 URL

基于GWO-LSTM与LSSVM的锂离子电池荷电状态与容量联合估计

Co-estimation of state of charge and capacity of lithium-ion battery based on GWO optimized LSTM and LSSVM

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 22

相关文章 15

编辑推荐

Metrics

本文评价

期刊信息

在线期刊

作者中心

审稿中心

联系我们

[1]	魏孟, 王桥, 叶敏, 廉高棨, 徐信芯. 基于dropout-MC递归神经网络的锂电池剩余寿命预测[J]. 汽车安全与节能学报, 2022, 13(3): 541-549.
[2]	谢翌, 江迪生, 张扬军, 李伟, 杨瑞, 钱煜平. 新能源汽车锂离子电池组SOC-SOP联合估计算法[J]. 汽车安全与节能学报, 2022, 13(3): 580-589.
[3]	来鑫, 孟正, 韩雪冰, 郑岳久. 基于特征电压模型的锂离子电池容量估计与RUL预测[J]. 汽车安全与节能学报, 2022, 13(1): 194-201.
[4]	董鹏, 张剑波, 王震坡. 基于电化学阻抗谱的锂离子电池析锂检测方法[J]. 汽车安全与节能学报, 2021, 12(4): 570-579.
[5]	王泽霖, 胡启国. 基于改进GWO算法的永磁同步电机滑模MRAS控制[J]. 汽车安全与节能学报, 2021, 12(3): 386-394.
[6]	穆道斌, 谢慧琳, 吴伯荣. 锂离子电池固体电解质的研究与进展[J]. 汽车安全与节能学报, 2020, 11(4): 415-427.
[7]	徐晓明, 袁秋奇, 张扬军, 胡昊. 极耳侧加热条件下锂离子电池热失控的数值分析[J]. 汽车安全与节能学报, 2020, 11(3): 388-396.
[8]	周旋, 周萍, 郑岳久, 韩雪冰, 褚政宇, 刘金海, 杨映华, 薛钢. 锂离子电池宽温度区间无析锂快充策略[J]. 汽车安全与节能学报, 2020, 11(3): 397-405.
[9]	刘中孝，葛昊，李哲，张雅琨，张剑波. 锂离子电池低温循环老化的空间分布与特征[J]. JASE, 2019, 10(4): 502-510.
[10]	Gregory L. Plett. 锂离子电池荷电状态不同估算方法的综述及讨论（英文）[J]. JASE, 2019, 10(3): 249-272.
[11]	王峰，罗玉涛 . 基于电池寿命的复合储能系统参数优化及能量管理[J]. JASE, 2019, 10(2): 211-218.
[12]	田野，宋凯. 考虑电池特性的多模型Kalman 滤波SOC 估计[J]. JASE, 2018, 9(2): 223-230.
[13]	吴正国，张剑波，李哲，LIAW Bor Yann. 锂离子电池加速老化温度应力的滥用边界[J]. JASE, 2018, 9(01): 99-109.
[14]	刘雄飞,崔彬,刘俊杰,林勇. 基于安时法的锂离子电池 SOC 估计实时校正方法[J]. JASE, 2017, 08(02): 178-182.
[15]	郑岳久，许霜霜，张振东. 三元锂电池荷电状态估计的传感器误差影响[J]. JASE, 2017, 08(02): 198-204.