Battery State of Health Estimation from Discharge Voltage Segments Using an Artificial Neural Network

IF 5.3 3区工程技术 Q1 ENGINEERING, MANUFACTURING International Journal of Precision Engineering and Manufacturing-Green Technology Pub Date : 2024-03-23 DOI:10.1007/s40684-024-00602-2

{"title":"Battery State of Health Estimation from Discharge Voltage Segments Using an Artificial Neural Network","authors":"","doi":"10.1007/s40684-024-00602-2","DOIUrl":null,"url":null,"abstract":"<h3>Abstract</h3> <p>Battery state of health (SOH) estimation is imperative for preventive maintenance, replacement, and end-of-life prediction of lithium ion batteries. Herein, we introduce a data-driven approach to state of health (SOH) prediction for battery cells using a Deep Neural Network (DNN). Our DNN model, trained on short discharge curve segments, outperforms Multilayer Perceptron (MLP) and Support Vector Regression (SVR) models. The Mutual Information (MI) score guides the selection of voltage range and width for model training, reflecting nonlinear degradation characteristics. A transfer learning strategy is applied for outlier cells, initially training on normal cells and fine-tuning with outlier cells, resulting in improved SOH predictions, particularly at higher cycles. The study finds that increasing the segment width reduces SOH prediction error, with the smallest segment of 0.05 V demonstrating good performance (RMSE of 0.0246), decreasing to 0.0142 at a width of 0.2 V. For outlier cells, transfer learning leads to a 48% reduction in RMSE. The partial segment-based approach offers potential for rapid SOH prediction in laboratory and field applications, enhancing efficiency in the development process.</p>","PeriodicalId":14238,"journal":{"name":"International Journal of Precision Engineering and Manufacturing-Green Technology","volume":"21 1","pages":""},"PeriodicalIF":5.3000,"publicationDate":"2024-03-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Precision Engineering and Manufacturing-Green Technology","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1007/s40684-024-00602-2","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MANUFACTURING","Score":null,"Total":0}

引用次数: 0

Abstract

Battery state of health (SOH) estimation is imperative for preventive maintenance, replacement, and end-of-life prediction of lithium ion batteries. Herein, we introduce a data-driven approach to state of health (SOH) prediction for battery cells using a Deep Neural Network (DNN). Our DNN model, trained on short discharge curve segments, outperforms Multilayer Perceptron (MLP) and Support Vector Regression (SVR) models. The Mutual Information (MI) score guides the selection of voltage range and width for model training, reflecting nonlinear degradation characteristics. A transfer learning strategy is applied for outlier cells, initially training on normal cells and fine-tuning with outlier cells, resulting in improved SOH predictions, particularly at higher cycles. The study finds that increasing the segment width reduces SOH prediction error, with the smallest segment of 0.05 V demonstrating good performance (RMSE of 0.0246), decreasing to 0.0142 at a width of 0.2 V. For outlier cells, transfer learning leads to a 48% reduction in RMSE. The partial segment-based approach offers potential for rapid SOH prediction in laboratory and field applications, enhancing efficiency in the development process.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

利用人工神经网络从放电电压分段估计电池健康状况

摘要电池健康状况（SOH）评估对于锂离子电池的预防性维护、更换和报废预测至关重要。在此，我们介绍了一种利用深度神经网络（DNN）进行电池健康状况（SOH）预测的数据驱动方法。我们的 DNN 模型在短放电曲线片段上经过训练，性能优于多层感知器 (MLP) 和支持向量回归 (SVR) 模型。互信息（MI）得分指导着模型训练中电压范围和宽度的选择，反映了非线性退化特性。针对离群电池采用了迁移学习策略，最初在正常电池上进行训练，然后利用离群电池进行微调，从而改进了 SOH 预测，尤其是在较高的周期下。研究发现，增加区段宽度可降低 SOH 预测误差，0.05 V 的最小区段表现良好（RMSE 为 0.0246），宽度为 0.2 V 时，误差降至 0.0142。对于离群单元，迁移学习可将 RMSE 降低 48%。基于部分区段的方法为实验室和现场应用中的快速 SOH 预测提供了潜力，提高了开发过程的效率。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

International Journal of Precision Engineering and Manufacturing-Green Technology ENGINEERING, MANUFACTURING-ENGINEERING, MECHANICAL

CiteScore

10.30

自引率

9.50%

发文量

审稿时长

5.3 months

期刊介绍： Green Technology aspects of precision engineering and manufacturing are becoming ever more important in current and future technologies. New knowledge in this field will aid in the advancement of various technologies that are needed to gain industrial competitiveness. To this end IJPEM - Green Technology aims to disseminate relevant developments and applied research works of high quality to the international community through efficient and rapid publication. IJPEM - Green Technology covers novel research contributions in all aspects of "Green" precision engineering and manufacturing.