AI-Integrated Experimental–Simulation Dataset for Predicting and Preventing Divergent Degradation in Parallel Battery Packs

Marlow et al. (2024) highlight how thermal gradients induce divergent degradation in parallel LIB packs, but there’s a lack of predictive, actionable tools for pack designers and operators. This idea proposes assembling a comprehensive dataset that fuses experimental measurements (impedance, current distribution, thermal imaging) with high-fidelity simulations (multi-physics, aging models) under varied pack configurations. Applying machine learning, the objective is to develop predictive models that not only forecast the onset of divergent degradation, but also suggest real-time interventions (e.g., dynamic cooling or current re-routing). Unlike current studies that treat experimental and simulation data separately, this integrated, AI-driven approach can reveal hidden predictors and nonlinear interactions, leading to smarter battery management systems and improved safety/longevity for large-scale energy storage.

References:

1. Degradation in parallel-connected lithium-ion battery packs under thermal gradients. Max Naylor Marlow, Jingyi Chen, Billy Wu (2024). Communications Engineer.