MS041 - Mathematical modeling, numerical simulations and AI techniques to enhance battery lifetime
Keywords: error analysis, foundation models for physical sciences, parametrized PDEs, PINNs, scientific machine learning
The existence of devices such as batteries to store energy reliably, is crucial both for the distribution and effective use of Renewable Energy sources and for recent applications in different technological areas (e.g. electric vehicles). Since the efficiency of a battery strongly depends on the charging ability, it is essential to develop highly green processes to control and reduce the distribution of material inside it, which compromises the battery durability. For this reason, enhancing the properties of such devices hardly depends on the ability of rationalizing charge-discharge processes and the distribution of metal growth at the electrodes.
This session aims to contribute to these goals by exploiting the synergy between numerical mathematics and the experimental world of batteries, with can benefit of mathematical modeling and AI techniques. We encourage an interdisciplinary approach along the following research topics.
Mathematical modeling and computational issues for:
- battery electrode morphology description (es. dendrites), for example phase-field modeling, pattern formation in PDEs, including stochastic operators;
- description and enhancement of charge-discharge cycling behavior;
- accurate and innovative numerical techniques for the models described above.
Deep Learning techniques for:
- comparison between simulations and experimental data (like time series describing the charging-recharging cycles and microscopy images of electrode degradation);
- identification of crucial model parameters in the PDE models;
- innovative numerical schemes, for example based on PINNs.
Contributions arising in other scientific issues, which may provide benefit to the above described context are more than welcome. This session is organized within the activities of the PRIN-PNRR project P20228C2PP BAT-MEN (BATtery Modeling, Experiments & Numerics) - Enhancing battery lifetime: mathematical modeling, numerical simulations and AI parameter estimation techniques for description and control of material localization processes.
This session aims to contribute to these goals by exploiting the synergy between numerical mathematics and the experimental world of batteries, with can benefit of mathematical modeling and AI techniques. We encourage an interdisciplinary approach along the following research topics.
Mathematical modeling and computational issues for:
- battery electrode morphology description (es. dendrites), for example phase-field modeling, pattern formation in PDEs, including stochastic operators;
- description and enhancement of charge-discharge cycling behavior;
- accurate and innovative numerical techniques for the models described above.
Deep Learning techniques for:
- comparison between simulations and experimental data (like time series describing the charging-recharging cycles and microscopy images of electrode degradation);
- identification of crucial model parameters in the PDE models;
- innovative numerical schemes, for example based on PINNs.
Contributions arising in other scientific issues, which may provide benefit to the above described context are more than welcome. This session is organized within the activities of the PRIN-PNRR project P20228C2PP BAT-MEN (BATtery Modeling, Experiments & Numerics) - Enhancing battery lifetime: mathematical modeling, numerical simulations and AI parameter estimation techniques for description and control of material localization processes.
