Technical University of Munich, School of Engineering and Design, Department of Energy and Process Engineering, Chair of Electrical Energy Storage Technology 80333 Munich, Germany
Auto-generated profile, based on publicly available information
Marcel Rogge
Technical University of Munich, School of Engineering and Design, Department of Energy and Process Engineering, Chair of Electrical Energy Storage Technology 80333 Munich, Germany
One of the most challenging tasks of modern battery management systems is the accurate state of health estimation. While physico-chemical models are accurate, they have high computational cost. Neural networks lack physical interpretability but are efficient. Physics-informed neural networks tackle the aforementioned shortcomings by combining the efficiency of neural networks with the accuracy of physico-chemical models. A physics-informed neural network is developed and evaluated against three different datasets: A pseudo-two-dimensional Newman model generates data at various state of health points. This dataset is fused with experimental data from laboratory measurements and vehicle field data to train a neural network in which it exploits correlation from internal modeled states to the measurable state of health. The resulting physics-informed neural network performs best with the synthetic dataset and achieves a root mean squared error below 2% at estimating the state of health. The root mean squared error stays within 3% for laboratory test data, with the lowest error observed for constant current discharge samples. The physics-informed neural network outperforms several other purely data-driven methods and proves its advantage. The inclusion of physico-chemical information from simulation increases accuracy and further enables broader application ranges.
Degradation models are important tools for understanding and mitigating lithium-ion battery aging, yet a universal model that can predict degradation under all operating conditions remains elusive. One challenge is the coupled influence of calendar and cycle aging phases on degradation mechanisms, such as solid electrolyte interphase (SEI) formation. In this work, we identify and systematically compare three different SEI interaction theories found in the literature, and apply them to experimental degradation data from a commercial lithium-ion cell. In a step-by-step process, and after careful data selection, we show that SEI delamination without any cracking of the active particles, and SEI microcracking, where cycling only affects SEI growth during the cycle itself, are both unlikely candidates. Instead, the results indicate that upon cycling, both the SEI and the active particle crack, and we provide a simple, 4-parameter equation that can predict the particle crack rate. Contrary to the widely-accepted Paris’ law, the particle crack rate decreases with increasing cycles, potentially due to changing intercalation dynamics resulting from the increasing surface-to-volume ratio of the active particles. The proposed model predicts SEI formation accurately at different storage conditions, while simply adding the degradation from pure calendar and cycle aging underestimates the total degradation.