Scholarlink

<p>State of health (SOH) is a key parameter of lithium batteries, and accurate prediction of SOH is essential for the healthy operation of battery systems. In this paper, macroscopic time and sample entropy are selected as parameters, and these two features are fused to form sample parameters as evaluation indicators of lithium battery health. In this paper, an improved probabilistic hierarchical simple particle swarm optimization (IPHSPSO) integrated with the support vector machine (SVM) is proposed to predict the life of lithium batteries. Enhanced speed and position adaptation schemes are introduced to enhance the global search ability of PSO. The proposed fusion model and improved PSO are studied on the public battery aging dataset of the Oxford University Battery Intelligence Laboratory. The comprehensive results of simulation experiments show that compared with other Oxford University lithium battery ICA peaks, discharge voltage integrals, battery capacity, microscopic time, and other feature fusion and single-action comparisons, as well as the traditional PSO-SVM and other particle swarm improved algorithms, the proposed sample entropy and macroscopic time feature fusion model and IPHSPSO have lower average error in predicting lithium battery SOH. Compared with the traditional single-feature prediction, feature fusion significantly improves the prediction accuracy of battery internal health characteristics (SOH). The proposed IPHSPSO-SVM model can accurately and effectively predict and judge the internal health status of lithium batteries.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Lithium battery health state prediction based on sample entropy and time feature fusion

  • 期刊论文

  • 发表日期：

    2025年7月19日

查看全文
  • Zedong Zhou,
  • Rui Zhong,
  • Yang Cao,
  • Xingbang Du,
  • Xinxin Guo

摘要

State of health (SOH) is a key parameter of lithium batteries, and accurate prediction of SOH is essential for the healthy operation of battery systems. In this paper, macroscopic time and sample entropy are selected as parameters, and these two features are fused to form sample parameters as evaluation indicators of lithium battery health. In this paper, an improved probabilistic hierarchical simple particle swarm optimization (IPHSPSO) integrated with the support vector machine (SVM) is proposed to predict the life of lithium batteries. Enhanced speed and position adaptation schemes are introduced to enhance the global search ability of PSO. The proposed fusion model and improved PSO are studied on the public battery aging dataset of the Oxford University Battery Intelligence Laboratory. The comprehensive results of simulation experiments show that compared with other Oxford University lithium battery ICA peaks, discharge voltage integrals, battery capacity, microscopic time, and other feature fusion and single-action comparisons, as well as the traditional PSO-SVM and other particle swarm improved algorithms, the proposed sample entropy and macroscopic time feature fusion model and IPHSPSO have lower average error in predicting lithium battery SOH. Compared with the traditional single-feature prediction, feature fusion significantly improves the prediction accuracy of battery internal health characteristics (SOH). The proposed IPHSPSO-SVM model can accurately and effectively predict and judge the internal health status of lithium batteries.