The performance of lithium-ion batteries(LIBs)gradually declines over time,making it critical to predict the battery’s state of health(SOH)in real-time.This paper presents a model that incorporates health indicators ...The performance of lithium-ion batteries(LIBs)gradually declines over time,making it critical to predict the battery’s state of health(SOH)in real-time.This paper presents a model that incorporates health indicators and ensemble Gaussian process regression(EGPR)to predict the SOH of LIBs.Firstly,the degradation process of an LIB is analyzed through indirect health indicators(HIs)derived from voltage and temperature during discharge.Next,the parameters in the EGPR model are optimized using the gannet optimization algorithm(GOA),and the EGPR is employed to estimate the SOH of LIBs.Finally,the proposed model is tested under various experimental scenarios and compared with other machine learning models.The effectiveness of EGPR model is demonstrated using the National Aeronautics and Space Administration(NASA)LIB.The root mean square error(RMSE)is maintained within 0.20%,and the mean absolute error(MAE)is below 0.16%,illustrating the proposed approach’s excellent predictive accuracy and wide applicability.展开更多
锂电池性能会随使用时间增加而逐步退化,若更换不及时,可能造成爆炸等严重事故。快速准确预测电池健康状态(state of health,SOH),对于锂电池系统管理和维护以及安全使用至关重要。本工作提出一种基于间接健康指标(health indicators,H...锂电池性能会随使用时间增加而逐步退化,若更换不及时,可能造成爆炸等严重事故。快速准确预测电池健康状态(state of health,SOH),对于锂电池系统管理和维护以及安全使用至关重要。本工作提出一种基于间接健康指标(health indicators,HIs)和高斯过程回归(Gaussian process regression,GPR)相结合预测锂电池SOH的机器学习模型。首先,通过分析锂电池放电过程,提取若干易于获得且适合动态操作的直接外部特征作为间接健康指标,并计算它们和SOH的相关性,最终筛选出平均放电电压、等压降放电时间、最高放电温度和平台期放电电压初始骤降值作为健康指标;其次,以上述健康指标作为输入特征,利用GPR算法建立锂电池退化模型,对NASA锂电池数据集进行预测,平均绝对误差(mean absolute error,MAE)不超过2%,均方根误差(root mean square error,RSME)控制在4%之内;最后,将本工作模型与其他常用机器学习模型进行比较,再将模型带入不同实验条件的电池中进行泛化性能分析,最大预测误差控制在6%之内,实验结果表明,本工作提出的间接健康指标和GPR模型具有相对较高的预测精度和优秀的泛化能力。展开更多
基金supported by Fundamental Research Program of Shanxi Province(No.202203021211088)Shanxi Provincial Natural Science Foundation(No.202204021301049).
文摘The performance of lithium-ion batteries(LIBs)gradually declines over time,making it critical to predict the battery’s state of health(SOH)in real-time.This paper presents a model that incorporates health indicators and ensemble Gaussian process regression(EGPR)to predict the SOH of LIBs.Firstly,the degradation process of an LIB is analyzed through indirect health indicators(HIs)derived from voltage and temperature during discharge.Next,the parameters in the EGPR model are optimized using the gannet optimization algorithm(GOA),and the EGPR is employed to estimate the SOH of LIBs.Finally,the proposed model is tested under various experimental scenarios and compared with other machine learning models.The effectiveness of EGPR model is demonstrated using the National Aeronautics and Space Administration(NASA)LIB.The root mean square error(RMSE)is maintained within 0.20%,and the mean absolute error(MAE)is below 0.16%,illustrating the proposed approach’s excellent predictive accuracy and wide applicability.
文摘锂电池性能会随使用时间增加而逐步退化,若更换不及时,可能造成爆炸等严重事故。快速准确预测电池健康状态(state of health,SOH),对于锂电池系统管理和维护以及安全使用至关重要。本工作提出一种基于间接健康指标(health indicators,HIs)和高斯过程回归(Gaussian process regression,GPR)相结合预测锂电池SOH的机器学习模型。首先,通过分析锂电池放电过程,提取若干易于获得且适合动态操作的直接外部特征作为间接健康指标,并计算它们和SOH的相关性,最终筛选出平均放电电压、等压降放电时间、最高放电温度和平台期放电电压初始骤降值作为健康指标;其次,以上述健康指标作为输入特征,利用GPR算法建立锂电池退化模型,对NASA锂电池数据集进行预测,平均绝对误差(mean absolute error,MAE)不超过2%,均方根误差(root mean square error,RSME)控制在4%之内;最后,将本工作模型与其他常用机器学习模型进行比较,再将模型带入不同实验条件的电池中进行泛化性能分析,最大预测误差控制在6%之内,实验结果表明,本工作提出的间接健康指标和GPR模型具有相对较高的预测精度和优秀的泛化能力。