储能科学与技术2025，Vol.14Issue(4)：1603-1616,14.DOI:10.19799/j.cnki.2095-4239.2024.0990
改进Sine混沌映射CO-ELM锂离子电池RUL预测
Remaining useful life prediction of a lithium-ion battery based on a cheetah optimization-extreme learning machine with improved Sine chaotic mapping
摘要
Abstract
To address the challenges of unstable predictions and accuracy when using extreme learning machines(ELMs)to predict the remaining useful life of lithium-ion batteries,this study proposes a cheetah optimization(CO)algorithm to optimize the ELM model performance.The equal voltage drop discharge time,extracted from the lithium-ion battery dataset,is employed as an indirect health factor.Furthermore,the CO algorithm is introduced to optimize the ELM parameters.This initial population of the CO algorithm is improved using sine chaotic mapping.The effectiveness and accuracy of the proposed model are verified using the battery dataset provided by the NASA Center for Excellence Prediction and the Oxford Battery Degradation Dataset from Oxford University.The optimal amount of training data and the ideal number of neurons are obtained through multiple experiments with the original ELM model.The residual service life of batteries is predicted using the proposed SCO-ELM model.Compared with the original ELM and the genetic algorithm-optimized ELM model,the proposed SCO-ELM model achieves a root mean square error below 0.004 and significantly faster prediction times.The prediction accuracy improves by 40%on average and the prediction speed is improved by more than 78%.Using the training results of battery B0005 to predict the performance of similar battery packs,the prediction accuracy improves by 25%on average,and the prediction speed increases by more than 75%.Thus,the experimental results confirm that the proposed method offers high prediction accuracy,fast computation speed,low operation complexity,and a stable model.关键词
锂离子电池/剩余使用寿命/极限学习机/猎豹优化/混沌映射Key words
lithium-ion battery/remaining useful life/extreme learning machine/cheetah optimization/chaotic mapping分类
信息技术与安全科学引用本文复制引用
王鹏,周俊,伍星,刘韬..改进Sine混沌映射CO-ELM锂离子电池RUL预测[J].储能科学与技术,2025,14(4):1603-1616,14.基金项目
云南省基础研究计划面上项目(202301AT070439),中船集团科技项目资助(2023530103002915). （202301AT070439）
