Scholarlink

<p>Accurate estimation of SOC is crucial to ensure electric vehicle (EV) battery safety, reliability, and performance. Conventional Coulomb Counting is one of the most widely used methods due to its simplicity and real-time capability. However, the method’s accuracy deteriorates over the long term of operation, due to numerical integration errors, sensor noise, and unmodeled capacity degradation. This paper proposes a trapezoidal numerical integration-based SOC estimation technique that compensates for capacity fade to lower cumulative drift at a relatively low cost. To assess and compare the conventional rectangular method of Euler with the proposed method, a synthetic drive cycle of 240&#xa0;min with a class interval for discharge and regenerative charging phases has been simulated. Reference SOC is obtained from ideal integration; the error metrics and performance indicators, including MAE, SOC drift, and ΔSOC distributions, are evaluated. The improved method consistently exhibits lower drift and error spread, especially during current polarity transitions. The proposed advancement provides a lightweight alternative to heavy model-based estimators and shows suitability for low-power battery management system applications in real EVs.</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

Enhancing SOC accuracy in electric vehicle batteries via trapezoidal integration and capacity degradation compensation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月31日

查看全文
  • Shreeram V. Kulkarni,
  • Sandeep Gupta,
  • G. Arjun,
  • Sombir Kundu,
  • Anand Shukla

摘要

Accurate estimation of SOC is crucial to ensure electric vehicle (EV) battery safety, reliability, and performance. Conventional Coulomb Counting is one of the most widely used methods due to its simplicity and real-time capability. However, the method’s accuracy deteriorates over the long term of operation, due to numerical integration errors, sensor noise, and unmodeled capacity degradation. This paper proposes a trapezoidal numerical integration-based SOC estimation technique that compensates for capacity fade to lower cumulative drift at a relatively low cost. To assess and compare the conventional rectangular method of Euler with the proposed method, a synthetic drive cycle of 240 min with a class interval for discharge and regenerative charging phases has been simulated. Reference SOC is obtained from ideal integration; the error metrics and performance indicators, including MAE, SOC drift, and ΔSOC distributions, are evaluated. The improved method consistently exhibits lower drift and error spread, especially during current polarity transitions. The proposed advancement provides a lightweight alternative to heavy model-based estimators and shows suitability for low-power battery management system applications in real EVs.