电工技术学报  2021, Vol. 36 Issue (24): 5176-5185    DOI: 10.19595/j.cnki.1000-6753.tces.201653
“新能源汽车电驱动系统与充放电技术”专题(特约主编:崔淑梅 教授 程 远 教授) |
基于LightGBM的电动汽车行驶工况下电池剩余使用寿命预测
肖迁1, 焦志鹏2, 穆云飞1, 陆文标1, 贾宏杰1
1.天津大学智能电网教育部重点实验室 天津 300072;
2.省部共建电工装备可靠性与智能化国家重点实验室(河北工业大学) 天津 300130
LightGBM Based Remaining Useful Life Prediction of Electric Vehicle Lithium-Ion Battery under Driving Conditions
Xiao Qian1, Jiao Zhipeng2, Mu Yunfei1, Lu Wenbiao1, Jia Hongjie1
1. Key Laboratory of Smart Grid of Ministry of Education Tianjin University Tianjin 300072 China;
2. State Key Laboratory of Reliability and Intelligence of Electrical Equipment Hebei University of Technology Tianjin 300130 China
全文: PDF (4868 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 行驶工况下电动汽车锂离子电池剩余使用寿命(RUL)衰退情况复杂,准确的RUL预测可为电池的定期维护和安全稳定运行提供指导,避免安全隐患。为此,该文提出一种适用于行驶工况下电动汽车电池的RUL预测方法。首先,针对行驶工况,提出一种基于轻量型梯度提升机(LightGBM)的RUL预测模型,利用元学习超参数优化方法对其进行超参数调优;其次,搭建行驶工况下电池全生命周期容量测试系统,模拟行驶工况下电池所受振动应力、充放电应力环境和测试电池容量衰退情况;然后,基于动态时间规整对容量衰退的相似性分析结果,使用生成对抗网络(GAN)生成新的容量序列;最后,通过实验数据验证所提模型和生成容量序列的有效性。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
肖迁
焦志鹏
穆云飞
陆文标
贾宏杰
关键词 电动汽车行驶工况锂离子电池剩余使用寿命轻量型梯度提升机    
Abstract:The degradation of the remaining useful life (RUL) for EV lithium-ion battery under driving conditions is complicated. The appropriate prediction of RUL can provide guidance for the periodic maintenance and stable operation to avoid the risks. Therefore, a RUL prediction method for driving conditions is proposed in this paper. Firstly, a light gradient boosting machine (LightGBM) based RUL prediction model is constructed, and the coefficients are obtained by the hyper parameter optimization (Hyperopt). Secondly, the experimental bench of battery cycle life capacity is established to simulate the vibration stress and charge-discharge stress, and the RUL degradation of battery under driving conditions is measured. Then, based on the dynamic time warping (DTW), the similarity of RUL degradation between driving conditions and static conditions is analyzed, and a new capacity sequence can be generated by the generative adversarial networks (GAN). Finally, experimental results verify the effectiveness of the proposed model and the generated capacity sequence.
Key wordsElectric vehicle    driving conditions    lithium-ion battery    remaining useful life    light gradient boosting machine (LightGBM)   
收稿日期: 2020-12-17     
PACS: TM911  
基金资助:国家自然科学基金(U2066213, 52107121)和中国博士后科学基金(2020M680880)资助项目
通讯作者: 焦志鹏 男,1993年生,硕士,研究方向为电动汽车储能技术。E-mail: 18722518050@163.com   
作者简介: 肖 迁 男,1988年生,博士,讲师,博士生导师,研究方向为分布式能源与微电网、直流配电网、电力电子技术及其在智能电网和综合能源系统中的应用、电池储能系统。E-mail: xiaoqian@tju.edu.cn
引用本文:   
肖迁, 焦志鹏, 穆云飞, 陆文标, 贾宏杰. 基于LightGBM的电动汽车行驶工况下电池剩余使用寿命预测[J]. 电工技术学报, 2021, 36(24): 5176-5185. Xiao Qian, Jiao Zhipeng, Mu Yunfei, Lu Wenbiao, Jia Hongjie. LightGBM Based Remaining Useful Life Prediction of Electric Vehicle Lithium-Ion Battery under Driving Conditions. Transactions of China Electrotechnical Society, 2021, 36(24): 5176-5185.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.201653          https://dgjsxb.ces-transaction.com/CN/Y2021/V36/I24/5176