Evaluation of Lithium-Ion Cell Characterization Procedures and Model Calibration Issues

The optimization of the characterization procedures needed for developing accurate models for lithium-ion batteries while reducing the overall testing and calibration times still represents a crucial aspect to be investigated. Indeed, despite the main standards and testing manuals provide basic guidelines for executing static and dynamic tests, various approaches are described in literature that demonstrate improvements in the model accuracy by adopting different testing parameters, leading to a consequent lack of standardization and uncertainties in the identification of the correct characterization procedure. With the aim of evaluating the model performances achievable by selecting specific testing parameters, this paper proposes a quantitative analysis that demonstrates their impact on the model accuracy. Different equivalent circuit models for a 90-Ah prismatic cell are developed based on basic pulsed discharge tests carried out with different testing parameters in terms of State of Charge (SoC) steps and rest times. The accuracy of the models is then validated by considering static, dynamic and custom current profiles. The results provide guidance to potential model calibration issues and contribute to the optimization of the testing time for the characterization procedures, highlighting that small SoC steps and long rest times are not strictly necessary for achieving accurate models.