基于SVR和电化学阻抗谱的锂电池内部温度在线估计  

Online estimation of internal temperature of lithium battery based on SVR and electrochemical impedance spectroscopy

作　　者：李强[1] 杨林[1] 李超凡 赵小巍 张树梅 LI Qiang;YANG Lin;LI Chaofan;ZHAO Xiaowei;ZHANG Shumei(School of Mechanical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China;SAIC Motor Corporation Limited,Shanghai 201804,China)

机构地区：[1]上海交通大学机械与动力工程学院,上海200240 [2]上汽集团研发总院,上海201804

出　　处：《电源技术》2024年第9期1738-1746,共9页Chinese Journal of Power Sources

基　　金：上海汽车工业科技发展基金会资助项目(2308);山东省重点研发计划(2023CXGC010210)。

摘　　要：准确实时地监测锂电池内部温度对于预防电池热失控至关重要。然而,目前尚缺乏有效的在线监测电池内部温度的方法。基于小型化阻抗测试系统,对锂离子电池在不同温度和荷电状态(SOC)下进行阻抗测试实验,研究电池温度和SOC对阻抗的影响,寻找与温度强相关而与SOC弱相关的特征频率。在此基础上,提出了一种基于支持向量回归(SVR)算法的锂电池内部温度估计算法,无需额外传感器,实现对电池内部温度的无损准确估计。Accurately and real-time monitoring the internal temperature of lithium batteries is crucial for preventing thermal runaway.However,there is currently no effective method for online monitoring of the internal temperature of batteries.Therefore,in this study,a miniaturized impedance testing system is utilized to perform impedance testing experiments on lithium ion batteries at various temperatures and state of charge(SOC)levels.The influence of battery temperature and SOC on impedance is investigated,aiming to identify characteristic frequencies strongly correlated with temperature and weakly correlated with SOC.On this basis,a lithium battery internal temperature estimation algorithm is proposed using the Support Vector Regression(SVR)technique,enabling nondestructive and accurate estimation of the internal temperature without the need for additional sensors.

关 键 词：锂电池 内部温度 电化学阻抗谱 支持向量回归 

分 类 号：TM912[电气工程—电力电子与电力传动]