基于充电片段和PSO-BP的锂电池SOH在线估计方法  

An online SOH estimation method for lithium batteries based on charging fragments and PSO-BP

作　　者：何山 赵宇明 田爱娜 姜久春 HE Shan;ZHAO Yuming;TIAN Aina;JIANG Jiuchun(Shenzhen Power Supply Bureau Co.,Ltd.,Shenzhen Guangdong 518000,China;Shenzhen Automotive Research Institute,Beijing Institute of Technology,Shenzhen Guangdong 518000,China;School of Electrical and Electronic Engineering,Hubei University of Technology,Wuhan Hubei 430068,China)

机构地区：[1]深圳供电局有限公司,广东深圳518000 [2]北京理工大学深圳汽车研究院,广东深圳518000 [3]湖北工业大学电气与电子工程学院,湖北武汉430068

出　　处：《电源技术》2025年第2期383-389,共7页Chinese Journal of Power Sources

基　　金：中国南方电网有限责任公司创新项目(090000KK52222138)。

摘　　要：由于锂离子电池具有自放电率低、比能量大等优点,目前常被应用于动力系统中。但由于电池老化过程中内部反应过于复杂,具有非线性、强耦合等特性,且健康状态不能直接测量,因此准确估算电池健康状态较难,如何准确对电池健康状态估算成为了电池领域内的研究热点。通过分析牛津大学实验室老化数据集,对温度和电压相关参数进行分析,发现随着循环的进行,温度变化率的斜率和等压升时间间隔变化的规律与容量的变换规律相同或者相反,因此提取温度和电压相关的参数作为健康特征。设计了一种基于粒子群优化-反向传播算法(PSO-BP)神经网络的电池健康状态估计模型,结果表明误差较小,在线估算误差能稳定在4%以内。Due to the advantages of low self-discharge rate and high specific energy,the lithium-ion batteries are commonly used in power systems.However,due to the complex internal reactions during the battery aging process,which exhibit nonlinear and strong coupling characteristics,the state of healthy cannot be directly measured,so it is difficult to accurately estimate the state of healthy.The accurate estimation of the state of healthy of batteries has become a hot research topic in the field of batteries.The Oxford aging dataset was analyzed,and the relevant parameters of temperature and voltage were examined.It’s found that the slope of the temperature change rate and the change of the interval of isothermal voltage rise with cycles are the same or opposite to the change of capacity.The temperature-related and voltage-related parameters were extracted as the health features.An estimation model for the battery state of healthy based on PSO-BP neural network was designed.The results show that the error is relatively small,and the online estimation error is stable within 4%.

关 键 词：锂离子电池 健康状态 PSO-BP神经网络 

分 类 号：TM912.9[电气工程—电力电子与电力传动]