Uncertainty-aware and explainable machine learning for early prediction of battery degradation trajectory
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Uncertainty-aware and explainable machine learning for early prediction of battery degradation trajectory. / Rieger, Laura Hannemose; Flores, Eibar; Nielsen, Kristian Frellesen; Norby, Poul; Ayerbe, Elixabete; Winther, Ole; Vegge, Tejs; Bhowmik, Arghya.
In: Digital Discovery, Vol. 2, No. 1, 2023, p. 112-122.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Uncertainty-aware and explainable machine learning for early prediction of battery degradation trajectory
AU - Rieger, Laura Hannemose
AU - Flores, Eibar
AU - Nielsen, Kristian Frellesen
AU - Norby, Poul
AU - Ayerbe, Elixabete
AU - Winther, Ole
AU - Vegge, Tejs
AU - Bhowmik, Arghya
N1 - Publisher Copyright: © 2023 The Author(s). Published by the Royal Society of Chemistry.
PY - 2023
Y1 - 2023
N2 - Enhancing cell lifetime is a vital criterion in battery design and development. Because lifetime evaluation requires prolonged cycling experiments, early prediction of cell aging can significantly accelerate both the autonomous discovery of better battery chemistries and their development into production. We demonstrate an early prediction model with reliable uncertainty estimates, which utilizes an arbitrary number of initial cycles to predict the whole battery degradation trajectory. Our autoregressive model achieves an RMSE of 106 cycles and a MAPE of 10.6% when predicting the cell's end of life (EOL). Beyond being a black box, we show evidence through an explainability analysis that our deep model learns the interplay between multiple cell degradation mechanisms. The learned patterns align with existing chemical insights into the rationale for early EOL despite not being trained for this or having received prior chemical knowledge. Our model will enable accelerated battery development via uncertainty-guided truncation of cell cycle experiments once the predictions are reliable.
AB - Enhancing cell lifetime is a vital criterion in battery design and development. Because lifetime evaluation requires prolonged cycling experiments, early prediction of cell aging can significantly accelerate both the autonomous discovery of better battery chemistries and their development into production. We demonstrate an early prediction model with reliable uncertainty estimates, which utilizes an arbitrary number of initial cycles to predict the whole battery degradation trajectory. Our autoregressive model achieves an RMSE of 106 cycles and a MAPE of 10.6% when predicting the cell's end of life (EOL). Beyond being a black box, we show evidence through an explainability analysis that our deep model learns the interplay between multiple cell degradation mechanisms. The learned patterns align with existing chemical insights into the rationale for early EOL despite not being trained for this or having received prior chemical knowledge. Our model will enable accelerated battery development via uncertainty-guided truncation of cell cycle experiments once the predictions are reliable.
U2 - 10.1039/d2dd00067a
DO - 10.1039/d2dd00067a
M3 - Journal article
AN - SCOPUS:85152765251
VL - 2
SP - 112
EP - 122
JO - Digital Discovery
JF - Digital Discovery
SN - 2635-098X
IS - 1
ER -
ID: 365812970