电源学报2026，Vol.24Issue(5)：256-264,9.DOI:10.13234/j.issn.2095-2805.2026.5.256
基于WPT与SSA优化的数据驱动电池健康状态估计
Data-driven Battery Health State Estimation Based on WPT and SSA Optimization
摘要
Abstract
The state of health(SOH)of Li-ion batteries gradually decreases with the number of charges and discharges,whose features are hidden in the physical information such as battery current and voltage.A data-driven method based on wavelet packet transform(WPT)and sparrow search algorithm(SSA)is proposed for online estimation of SOH,and the feature indices of discharge voltage plateau time(DVPT)and constant current charging time(CCCT)are proposed.The algorithm is verified on the University of Maryland battery public dataset.The results show that the proposed DVPT characteristic metric has a strong correlation with SOH,with an average Pearson correlation coefficient of 98.38%across the four batteries;the model predicted the optimal results with RMSE of 0.019 1,MAE of 0.012 5,and R2 of 97.78%.关键词
锂电池/健康状态/数据驱动/麻雀搜索算法Key words
Lithium battery/state of health/data-driven/sparrow search algorithm分类
信息技术与安全科学引用本文复制引用
黄杰明,黄小荣,张庆波,林炜,吴树平,罗俊杰..基于WPT与SSA优化的数据驱动电池健康状态估计[J].电源学报,2026,24(5):256-264,9.基金项目
南方电网公司科技资助项目(031900KK52220011)This work is supported by Science and Technology Program of China Southern Power Grid under the grant 031900KK52220011 （031900KK52220011）
