Scholarlink

Lithium-ion batteries are widely used in various electronic devices and electric vehicles, and accurately estimating their state of health (SOH) is crucial for ensuring performance and longevity. This study introduces a novel approach for SOH estimation in lithium-ion batteries, employing a method that integrates multiple health feature extraction with Gaussian process regression (GPR) optimized by the grey wolf optimizer (GWO). By extracting multi-dimensional health features, including voltage, time, and temperature, and optimizing the hyperparameters of the Gaussian Process Regression (GPR) model using the Grey Wolf Optimizer (GWO) algorithm, a high-precision and robust state-of-health (SOH) estimation model was constructed. The experimental results demonstrate that this method exhibits a significant enhancement in both estimation accuracy and stability compared to conventional algorithms, thereby providing effective support for the reliable management of lithium-ion batteries.

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

SOH Estimation of Power Lithium-Ion Batteries Based on Gaussian Process Regression Model Optimized by Multi-feature Extraction and Grey Wolf Optimization

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Jiayin Xu,
  • Shaohua Chen,
  • Zihe Liu,
  • Xiaojie Xu

摘要

Lithium-ion batteries are widely used in various electronic devices and electric vehicles, and accurately estimating their state of health (SOH) is crucial for ensuring performance and longevity. This study introduces a novel approach for SOH estimation in lithium-ion batteries, employing a method that integrates multiple health feature extraction with Gaussian process regression (GPR) optimized by the grey wolf optimizer (GWO). By extracting multi-dimensional health features, including voltage, time, and temperature, and optimizing the hyperparameters of the Gaussian Process Regression (GPR) model using the Grey Wolf Optimizer (GWO) algorithm, a high-precision and robust state-of-health (SOH) estimation model was constructed. The experimental results demonstrate that this method exhibits a significant enhancement in both estimation accuracy and stability compared to conventional algorithms, thereby providing effective support for the reliable management of lithium-ion batteries.