Scholarlink

<p>The internal polarization phenomena observed in lithium-ion batteries used in the battery mining vehicles is exacerbated by high pulse discharging and frequency cycling, along with over-discharge conditions. This performance decline restricts the accuracy of state of charge (SOC) estimation under long-term climbing conditions with over-discharge. In order to improve the accuracy and robustness of the SOC estimation of lithium-ion batteries for the battery mining vehicles, a closed-loop method based on machine learning and Kalman filtering is combined. Firstly, the long short-term memory (LSTM) network is utilized to establish the offline training model. Then, the Sage–Husa adaptive Kalman filter (AKF) is introduced to obtain the online estimation. The observed data from lithium-ion batteries for the battery mining vehicles has been selected to verify the accuracy and robustness of combined method under over-discharge conditions. Compared with traditional methods, the closed-loop SOC estimation method can achieve higher accuracy and strong robustness with mean absolute error and root-mean-squared error below 1%.</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

Closed-Loop State of Charge Estimation of Lithium-Ion Batteries Based on Machine Learning and Kalman Filtering

  • 期刊论文

  • 发表日期：

    2025年5月18日

查看全文
  • Chuanwei Zhang,
  • Ting Wang,
  • Binrui Zhang,
  • Lin Qiao,
  • Gaoqi Lian,
  • Meng Wei

摘要

The internal polarization phenomena observed in lithium-ion batteries used in the battery mining vehicles is exacerbated by high pulse discharging and frequency cycling, along with over-discharge conditions. This performance decline restricts the accuracy of state of charge (SOC) estimation under long-term climbing conditions with over-discharge. In order to improve the accuracy and robustness of the SOC estimation of lithium-ion batteries for the battery mining vehicles, a closed-loop method based on machine learning and Kalman filtering is combined. Firstly, the long short-term memory (LSTM) network is utilized to establish the offline training model. Then, the Sage–Husa adaptive Kalman filter (AKF) is introduced to obtain the online estimation. The observed data from lithium-ion batteries for the battery mining vehicles has been selected to verify the accuracy and robustness of combined method under over-discharge conditions. Compared with traditional methods, the closed-loop SOC estimation method can achieve higher accuracy and strong robustness with mean absolute error and root-mean-squared error below 1%.