Battery calendar aging and machine learning Comment

Article Chemistry, Physical

Feature engineering for machine learning enabled early prediction of battery lifetime

Noah H. Paulson et al.

Summary: This study investigates the application of machine learning in battery lifetime prediction using a dataset of cycling data from various cathode chemistries. The results show that machine learning tools can accurately predict lithium-ion battery lifetimes even with limited preliminary cycles. Additionally, the study explores feature selection strategies and demonstrates the generalization of machine learning models across different battery chemistries.

JOURNAL OF POWER SOURCES (2022)

Add to Collection

Article Electrochemistry

Challenging Practices of Algebraic Battery Life Models through Statistical Validation and Model Identification via Machine-Learning

Paul Gasper et al.

Summary: The article introduces a novel procedure for automatically identifying reduced-order degradation models, which are statistically validated using cross-validation, sensitivity analysis, and uncertainty quantification via bootstrapping, demonstrating that previously known physical relationships can be empirically rediscovered using machine learning.

JOURNAL OF THE ELECTROCHEMICAL SOCIETY (2021)

Add to Collection

Article Energy & Fuels

Corrosion of lithium metal anodes during calendar ageing and its microscopic origins

David T. Boyle et al.

Summary: Rechargeable lithium metal batteries must have a long cycle life and calendar life. The loss of capacity during calendar ageing is caused by chemical corrosion of Li and the continuous growth of the solid electrolyte interphase. Functional electrolytes must minimize the rate of solid electrolyte interphase growth and the surface area of electrodeposited Li metal to ensure long battery life.

NATURE ENERGY (2021)

Add to Collection

Article Energy & Fuels

Calendar aging of silicon-containing batteries

Josefine D. McBrayer et al.

Summary: Silicon-containing batteries are gaining popularity in the mass market, but the calendar aging behaviors of silicon anodes have not been well studied. Researchers are discussing the challenges surrounding the long-term stability of silicon anodes for lithium-ion batteries, especially in high-energy batteries for automotive applications. Future research should focus on assessing and mitigating the time-dependent degradation of silicon-containing batteries to fully realize their benefits.

NATURE ENERGY (2021)

Add to Collection

Article Chemistry, Multidisciplinary

A machine learning framework for early detection of lithium plating combining multiple physics-based electrochemical signatures

Bor-Rong Chen et al.

Summary: The research established a framework for early detection of loss of Li inventory in lithium-ion batteries, distinguishing between Li plating and SEI-formation-dominated cells, using machine learning and multiple electrochemical signatures to improve classification accuracy.

CELL REPORTS PHYSICAL SCIENCE (2021)

Add to Collection

Article Electrochemistry

Early battery performance prediction for mixed use charging profiles using hierarchal machine learning

M. Ross Kunz et al.

Summary: Using data analysis methods such as machine learning can accelerate the battery performance validation process, and it is possible to predict battery performance using only a small amount of cycling data during training, but relying solely on cycling data for battery lifetime analysis may have limitations.

BATTERIES & SUPERCAPS (2021)

Add to Collection

Article Energy & Fuels