Scholarlink

<p>Lithium Iron Phosphate (LFP) batteries exhibit pronounced hysteresis phenomena, where electrochemical characteristics differ significantly between charge and discharge processes. This hysteresis effect causes different Open Circuit Voltage (OCV) values at identical State of Charge (SOC) levels depending on charge/discharge history, which complicates accurate battery modeling. Existing battery models incorporating hysteresis fail to adaptively capture dynamic behavior that varies with charge/discharge rates (C-rate) and rest periods. This study extends the One-State Hysteresis Model (OSHM), an existing equivalent circuit model that accounts for hysteresis, to address these limitations. The proposed model introduces a scaling parameter α to capture C-rate dependent hysteresis magnitude and employs a relaxation time constant τ<sub>rest</sub> to represent voltage relaxation during rest periods. Model validation is performed using railway vehicle drive cycles with experimental data at various C-rates. Validation results demonstrate 15% reduction in terminal voltage Root Mean Square Error (RMSE) compared to 2RC models, 1.6% reduction compared to OSHM, and 26% improvement in prediction consistency (Interquartile Range, IQR) over OSHM under dynamic operating conditions.</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

Dynamic Hysteresis Modeling for LFP Batteries: Adaptive State Diagnosis Under Variable Operating Conditions

  • 期刊论文

  • 发表日期：

    2026年3月2日

查看全文
  • Jae-Uk Sim,
  • Seon-Woong Kim,
  • In-Ho Cho

摘要

Lithium Iron Phosphate (LFP) batteries exhibit pronounced hysteresis phenomena, where electrochemical characteristics differ significantly between charge and discharge processes. This hysteresis effect causes different Open Circuit Voltage (OCV) values at identical State of Charge (SOC) levels depending on charge/discharge history, which complicates accurate battery modeling. Existing battery models incorporating hysteresis fail to adaptively capture dynamic behavior that varies with charge/discharge rates (C-rate) and rest periods. This study extends the One-State Hysteresis Model (OSHM), an existing equivalent circuit model that accounts for hysteresis, to address these limitations. The proposed model introduces a scaling parameter α to capture C-rate dependent hysteresis magnitude and employs a relaxation time constant τrest to represent voltage relaxation during rest periods. Model validation is performed using railway vehicle drive cycles with experimental data at various C-rates. Validation results demonstrate 15% reduction in terminal voltage Root Mean Square Error (RMSE) compared to 2RC models, 1.6% reduction compared to OSHM, and 26% improvement in prediction consistency (Interquartile Range, IQR) over OSHM under dynamic operating conditions.