Modeling of Li-ion Battery Thermal Runaway: Insights into Modeling and Prediction

This article discusses the dual challenges of modeling and predicting thermal runaway (TR) in li-ion batteries. It explores the current challenges of TR modeling, the methods for progressing from single cells to battery packs, and future directions involving the use of probabilistic methods and ML/AI to tackle this multifaceted issue. Accurate prediction of TR events remains a major challenge, especially because of the cell-to-cell variability, but also due to parametrization, which requires complex experimentation or exhaustive experimental data for training data-driven ML models. Continued research and innovation in modeling and predictive techniques are essential for ensuring safer and more reliable battery systems.