LGApr 26, 2024

Optimizing Cycle Life Prediction of Lithium-ion Batteries via a Physics-Informed Model

arXiv:2404.17174v21 citationsh-index: 1Trans. Mach. Learn. Res.
Originality Incremental advance
AI Analysis

This work addresses battery performance prediction for technology development, offering an incremental improvement in interpretability and robustness.

The paper tackled predicting the cycle lifetime of lithium-ion batteries by introducing a hybrid physics-informed self-attention model that uses early-cycle data to reconstruct entire capacity loss curves, achieving comparable performance to existing models while providing more robustness and interpretability.

Accurately measuring the cycle lifetime of commercial lithium-ion batteries is crucial for performance and technology development. We introduce a novel hybrid approach combining a physics-based equation with a self-attention model to predict the cycle lifetimes of commercial lithium iron phosphate graphite cells via early-cycle data. After fitting capacity loss curves to this physics-based equation, we then use a self-attention layer to reconstruct entire battery capacity loss curves. Our model exhibits comparable performances to existing models while predicting more information: the entire capacity loss curve instead of cycle life. This provides more robustness and interpretability: our model does not need to be retrained for a different notion of end-of-life and is backed by physical intuition.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes