There is a rising need for accurate battery state of health (SOH) diagnosis in electric vehicle maintenance and second-life evaluation. However, existing methods suffer from the transition from cell-level tests to real-world vehicle applications due to the ignorance of incorporating laboratory tests with large-scale, time-varying field data. This paper proposes a framework combining the system-level capacity calculation and cell-level decoupling experiment for battery system capacity diagnosis. A modified regional capacity calculation method for online applications is presented, and the regional capacity of the battery under various temperatures and SOHs is experimentally determined to decouple various working conditions. This work highlights the opportunity to integrate laboratory test data to leverage unlabelled field data for capacity diagnosis while revealing the characteristics of battery capacity under different working conditions.
Keywords Lithium-ion batteries, electric vehicles, health diagnosis, field data, decoupling experiment