Quantifying battery stress factors: An intuitive generic model for calendar and cycle aging simulation using time series analysis

This study introduces a comprehensive and generic aging model that disentangles and quantifies the isolated effects of individual stress factors on lithium-ion battery degradation. Unlike traditional empirical or semi-empirical models, which often combine stress factors into aggregate predictions, the proposed framework leverages time series analysis to resolve the distinct contributions of temperature, state of charge (SOC), depth of discharge (DOD), and current rate (C-rate). This allows not only the capture of overall degradation trends but also the interpretable parameterization of stress-dependent contributions across the operating ranges of NMC cells. The model achieves high predictive accuracy, with a mean absolute error (MAE) of 1.10% for training and 0.61% for validation against data from realistic load profiles. Key findings reveal that the SOC plays a crucial role in battery aging, particularly at elevated levels, with diminishing negative effects observed beyond 85% SOC. In addition, DOD emerged as the most significant factor, with higher DOD levels correlating with accelerated aging, likely due to mechanical stresses impacting electrode integrity. The analysis showed that while SOC consistently contributes to aging, the C-rate exhibited minimal influence within the tested parameters. This study emphasizes the importance of understanding these aging factors for developing effective battery management strategies. Despite the model’s predictive performance, several limitations were identified, including challenges in representing extreme conditions and accurately modeling non-linear capacity fade. Understanding these limitations is essential for enhancing the model’s applicability and reliability in real-world scenarios. Overall, this approach advances beyond cumulative modeling frameworks by offering interpretable and transferable model parameters, directly linking operating conditions to aging factors and corresponding aging parameters. These insights not only enhance predictive capability but also support the design of optimized battery management strategies aimed at mitigating dominant aging conditions.