Scholarlink

<p>The accurate estimation of state-of-charge (SoC) is critical within smart battery management systems (BMS). Despite numerous research articles discussing effective SoC estimation techniques, there remains a need to enhance the accuracy of the SoC estimation module, especially given its significance in various vehicular applications. In this context, the article examines machine learning operations (MLOps) for SoC estimation in Li-ion batteries, focusing on principles and practices aimed at effectively managing and implementing machine learning models in practical scenarios. Operational machine learning for SoC estimation involves leveraging real-time data from battery systems to continually enhance the precision and reliability of SoC predictions. Furthermore, the article extensively discusses the requirements, operations, and constraints associated with the models. It specifically addresses the challenge of model selection in MLOps, taking into account critical aspects of SoC estimation and performance metrics of machine learning models. The article aims to offer clarity on selecting, utilizing, and feasibly implementing MLOps models for advancing SoC estimation toward potential real-time applications. The study utilized real-life data from Panasonic 18650PF Li-ion battery cells to train and test the MLOps models under consideration. The machine learning application was implemented using Python.</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

Operational machine learning based approach for effective state-of-charge estimation in Li-ion batteries

  • 期刊论文

  • 发表日期：

    2025年11月13日

查看全文
  • Tikam Bhardwaj,
  • Vijay Kale,
  • Makarand Sudhakar Ballal,
  • Sudarshan Khond

摘要

The accurate estimation of state-of-charge (SoC) is critical within smart battery management systems (BMS). Despite numerous research articles discussing effective SoC estimation techniques, there remains a need to enhance the accuracy of the SoC estimation module, especially given its significance in various vehicular applications. In this context, the article examines machine learning operations (MLOps) for SoC estimation in Li-ion batteries, focusing on principles and practices aimed at effectively managing and implementing machine learning models in practical scenarios. Operational machine learning for SoC estimation involves leveraging real-time data from battery systems to continually enhance the precision and reliability of SoC predictions. Furthermore, the article extensively discusses the requirements, operations, and constraints associated with the models. It specifically addresses the challenge of model selection in MLOps, taking into account critical aspects of SoC estimation and performance metrics of machine learning models. The article aims to offer clarity on selecting, utilizing, and feasibly implementing MLOps models for advancing SoC estimation toward potential real-time applications. The study utilized real-life data from Panasonic 18650PF Li-ion battery cells to train and test the MLOps models under consideration. The machine learning application was implemented using Python.