Advances in Battery Management: Extending State of Health Assessment Beyond Automotive Applications

Strategic Insights -

Researchers from Jilin University, in partnership with the State Grid Jilin Electric Power Company Limited, have made a significant breakthrough in battery management by refining the state of health (SOH) assessment of power batteries. Through a data-driven methodology, they have successfully extended the SOH assessment window to early stages and second-life applications, while reducing dependence on training data. This achievement has the potential to mitigate carbon costs and enhance lifecycle management.

Key Takeaways:

  • The researchers developed a composite aging dataset with fast-charging and low-temperature conditions to investigate feature effectiveness evolution across the battery lifecycle.
  • By integrating Transformer and Bi-directional Long Short-Term Memory (BiLSTM) architectures with transfer learning, the team identified optimal feature-model combinations and extended the SOH assessment window.
  • Experiments showed that in second-life applications, SOH estimation RMSE remained below 2%, while transfer learning enabled the valid assessment range to 30% SOH using only 15% target data.
  • The study concluded that this research provides a critical approach to reduce carbon costs and enhance lifecycle management.
  • The authors presented their findings in the peer-reviewed journal "Energy," published by Elsevier.
  • The research was partially funded by the State Grid Jilin Electric Power Company Limited through the "special cost project" grant.

Statistics:

  • The SOH estimation RMSE remained below 2% in second-life applications.
  • Transfer learning enabled the valid assessment range to 30% SOH using only 15% target data.
  • The study investigated feature effectiveness evolution across the battery lifecycle using a composite aging dataset.
  • The research provided a critical approach to reduce carbon costs and enhance lifecycle management.
  • The authors identified optimal feature-model combinations through the integration of Transformer and BiLSTM architectures with transfer learning.

Sources:

  • VerticalNews
  • A Data-driven Methodology for Early-stage Estimation and Second-life Applicability Assessment Toward Lifecycle Refined Management of Power Batteries. Energy, 2025;335. (Elsevier - www.elsevier.com; Energy - www.journals.elsevier.com/energy/)
  • Jilin University, State Key Lab Automot Chassis Integrat & Bion, Changchun 130022, People's Republic of China
  • State Grid Jilin Electric Power Company Limited