Search papers, labs, and topics across Lattice.
This study introduces a regime-aware, physics-guided framework that leverages a combination of temperature, voltage, force, deformation, and state-of-charge measurements to enhance early warning systems for lithium-ion battery thermal runaway. By employing a lightweight convolutional classifier to categorize mechanical signals into safe, warning, or danger regimes, the framework integrates these regime estimates into a causal temporal convolutional model for improved detection and time-to-disaster estimation. The results demonstrate a significant lead time improvement of 69.6% over the strongest baseline, achieving an F1 score of 0.89 and underscoring the critical role of mechanical precursors in early warning systems.
Mechanical signals can provide critical early warnings for lithium-ion battery thermal runaway, achieving a lead time that outpaces existing methods by nearly 70%.
Thermal runaway in lithium-ion batteries poses a major safety risk to electric vehicles and energy storage systems. Current early-warning methods depend mainly on temperature and may therefore miss mechanical precursors that emerge before rapid heating. We introduce a regime-aware, physics-guided framework that integrates temperature, voltage, force, deformation, and state-of-charge measurements for early warning under controlled mechanical abuse. A lightweight convolutional classifier first infers safe, warning, or danger regimes from mechanical signals. These regime estimates then condition a causal temporal convolutional backbone through feature-wise linear modulation, physics-biased attention, and regime-dependent gating. Joint learning unifies regime identification, thermal-runaway detection, and time-to-disaster estimation. We evaluate the framework using leave-one-experiment-out cross-validation on 30 mechanical-abuse tests across state-of-charge levels of 10%, 50%, and 90% and two loading protocols. The method achieves an F1 score of 0.89, a high-temperature prediction root-mean-square error of 12.3 {\deg}C, a mean warning lead time of 15.6 s, a detection success rate of 0.92, and an experiment-level false alarm rate of 2.7%. Its lead time exceeds that of the strongest baseline by 69.6%. Removing force reduces the lead time by 60.3%, highlighting the value of mechanical precursors. These results support regime-aware thermo-mechanical fusion as a promising strategy for earlier and more reliable thermal-runaway warning under controlled abuse conditions.
