Scholarlink

<p>Addressing the issue that the online estimation of state of charge and state of health of lithium-ion batteries is susceptible to temperature and noise influences, this paper proposes a collaborative estimation algorithm that combines intelligent algorithms with adaptive filtering. Based on the mechanism of temperature affecting capacity and open-circuit voltage, a temperature-adaptive dual-polarization equivalent circuit model is constructed. The finite memory multi-innovation recursive least squares algorithm is adopted for high-precision parameter identification. To address the time-varying noise problem, grey wolf optimization and a moving window function are introduced to construct an intelligent optimization adaptive extended Kalman filter, achieving co-estimation of SOC and SOH on dual time scales. The validation results under different temperatures and operating conditions show that the mean absolute errors of SOC and SOH estimations are below 0.75% and 0.90%, respectively, providing theoretical support for safe and efficient battery management.</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

Collaborative estimation of lithium-ion battery state based on multi-innovation identification and improved dual adaptive filtering

  • 期刊论文

  • 发表日期：

    2026年7月9日

查看全文
  • Lu Chen,
  • Sipeng Jiang,
  • Shunli Wang,
  • Lei Chen,
  • Carlos Fernandez

摘要

Addressing the issue that the online estimation of state of charge and state of health of lithium-ion batteries is susceptible to temperature and noise influences, this paper proposes a collaborative estimation algorithm that combines intelligent algorithms with adaptive filtering. Based on the mechanism of temperature affecting capacity and open-circuit voltage, a temperature-adaptive dual-polarization equivalent circuit model is constructed. The finite memory multi-innovation recursive least squares algorithm is adopted for high-precision parameter identification. To address the time-varying noise problem, grey wolf optimization and a moving window function are introduced to construct an intelligent optimization adaptive extended Kalman filter, achieving co-estimation of SOC and SOH on dual time scales. The validation results under different temperatures and operating conditions show that the mean absolute errors of SOC and SOH estimations are below 0.75% and 0.90%, respectively, providing theoretical support for safe and efficient battery management.