Scholarlink

<p>The accurate estimation of temperature states in lithium-ion batteries is crucial for the performance and reliability of battery management systems. Reliable temperature estimation plays a vital role in ensuring safe operation, extending service life, and preventing thermal runaway in batteries. However, existing studies exhibit deficiencies in parameter sensitivity and noise adaptability, which limit the accuracy and reliability of temperature estimations. In this paper, we propose a novel method for battery temperature state estimation that combines a polynomial approximate thermal model with an adaptive unscented Kalman filter (AUKF). To achieve accurate and efficient parameter identification, we introduce an enhanced parrot optimization algorithm (EPO). Building upon this, the AUKF is employed to estimate both the surface and core temperatures of the battery, thereby improving the accuracy and robustness of the estimations. Experimental results demonstrate the effectiveness of the proposed method. Compared to other optimization algorithms, the EPO shows enhanced speed and accuracy in parameter recognition. At various test conditions, including the federal urban driving schedule (FUDS), dynamic stress test (DST), and hybrid electric vehicle (HEV) scenarios, the maximum root mean square error (RMSE) for the core temperature estimate using AUKF is 0.2705 °C, while the maximum RMSE for the surface temperature estimate is 0.0308 °C. These results indicate that the proposed method offers good accuracy and adaptability and is low-cost, simple to implement, and suitable for practical applications.</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

Multi-condition temperature state estimation of lithium-ion battery based on enhanced parrot optimization and adaptive unscented Kalman filter

  • 期刊论文

  • 发表日期：

    2025年9月27日

查看全文
  • Yuhai Yao,
  • Jun Xie,
  • Xiaojian Ma,
  • Yixiao Zhang,
  • Yutong Zhang,
  • Yan Li,
  • Qing Xie,
  • Nan Wang

摘要

The accurate estimation of temperature states in lithium-ion batteries is crucial for the performance and reliability of battery management systems. Reliable temperature estimation plays a vital role in ensuring safe operation, extending service life, and preventing thermal runaway in batteries. However, existing studies exhibit deficiencies in parameter sensitivity and noise adaptability, which limit the accuracy and reliability of temperature estimations. In this paper, we propose a novel method for battery temperature state estimation that combines a polynomial approximate thermal model with an adaptive unscented Kalman filter (AUKF). To achieve accurate and efficient parameter identification, we introduce an enhanced parrot optimization algorithm (EPO). Building upon this, the AUKF is employed to estimate both the surface and core temperatures of the battery, thereby improving the accuracy and robustness of the estimations. Experimental results demonstrate the effectiveness of the proposed method. Compared to other optimization algorithms, the EPO shows enhanced speed and accuracy in parameter recognition. At various test conditions, including the federal urban driving schedule (FUDS), dynamic stress test (DST), and hybrid electric vehicle (HEV) scenarios, the maximum root mean square error (RMSE) for the core temperature estimate using AUKF is 0.2705 °C, while the maximum RMSE for the surface temperature estimate is 0.0308 °C. These results indicate that the proposed method offers good accuracy and adaptability and is low-cost, simple to implement, and suitable for practical applications.