Scholarlink

<p>Accurate estimation of state of charge (SOC) is a critical function of battery management system under complex working conditions. To improve the adaptability of SOC estimation to different working conditions, an error compensation strategy is proposed based on the extended single particle model (ESPM) in this paper, which combines backpropagation neural network (BPNN) with extended Kalman filter to compensate for the linearization error. The initial weights and thresholds of the BPNN are optimized by the animated oat optimization (AOO) algorithm, thereby addressing its tendency to fall into local optima. The SOC is preliminarily estimated based on ESPM, and then corrected by the error prediction results from the trained AOO-BPNN model. Under different conditions, the root mean square error (RMSE) of SOC estimation remains below 0.60%. Under different temperatures and aging cycles, the RMSEs do not exceed 0.45% and 0.31%, respectively. The research findings demonstrate that this strategy exhibits strong adaptability and robustness.</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

An error compensation strategy for state-of-charge estimation based on ESPM and AOO-BPNN

  • 期刊论文

  • 发表日期：

    2026年7月30日

查看全文
  • Chenhui Li,
  • Zhan Ma,
  • Zhongkai Zhou

摘要

Accurate estimation of state of charge (SOC) is a critical function of battery management system under complex working conditions. To improve the adaptability of SOC estimation to different working conditions, an error compensation strategy is proposed based on the extended single particle model (ESPM) in this paper, which combines backpropagation neural network (BPNN) with extended Kalman filter to compensate for the linearization error. The initial weights and thresholds of the BPNN are optimized by the animated oat optimization (AOO) algorithm, thereby addressing its tendency to fall into local optima. The SOC is preliminarily estimated based on ESPM, and then corrected by the error prediction results from the trained AOO-BPNN model. Under different conditions, the root mean square error (RMSE) of SOC estimation remains below 0.60%. Under different temperatures and aging cycles, the RMSEs do not exceed 0.45% and 0.31%, respectively. The research findings demonstrate that this strategy exhibits strong adaptability and robustness.