Multiscale Digital-Twin Modeling of Energy-Storage Materials by Integrating First-Principles Calculations and Machine-Learning Interatomic Potentials
DOI:
https://doi.org/10.4108/ew.14196Keywords:
first-principles calculations, machine-learning interatomic potential, equivalent graph neural network, active learning, multiscale digital twin, energy-storage materialsAbstract
Conventional modeling of energy-storage materials is constrained by the computational cost of first-principles calculations, limited electrochemical-state awareness in machine-learning interatomic potentials, and weak coupling between materials-scale models and online battery decision-making. This study proposes a first-principles-constrained electrochemically conditioned machine-learning interatomic potential (FP-EMLP) integrated with a multiscale digital-twin framework (FP-EMLP-DT). A multistate configuration space spanning ion concentration, temperature, defects, strain, and phase boundaries is combined with active learning based on committee uncertainty, structural novelty, and physics-based risk. The equivariant graph model incorporates electrochemical-state gating, charge-conserving projection, and differentiable long-range electrostatics, and is trained using energies, forces, stresses, equilibrium voltages, and migration barriers. Predicted diffusion coefficients, free energies, elastic tensors, and chemical expansion coefficients are transferred to an electrochemical-mechanical-thermal-damage model. The resulting model is further integrated with a physics-based reduced-order network, uncertainty-gated residual learning, online data assimilation, and degradation-aware control. FP-EMLP achieves mean absolute errors of 6.5 meV atom−1, 33 meV Å−1, 0.051 V, and 0.038 eV for energy, force, voltage, and migration barrier, respectively, with a 4.8% aggregate relative error across multiscale fields. Active learning reduces required first-principles labels by 33.3% while covering 91.2% of high-risk configurations. FP-EMLP-DT achieves mean absolute errors of 0.93% for state of charge and 1.21% for state of health. Degradation-aware control reduces maximum material stress and damage increment by 16.5% and 36.8%, respectively, demonstrating an effective multiscale pathway from electronic structure to online battery management.
Downloads
References
[1] Semeraro C, Aljaghoub H, Abdelkareem MA, et al. Guidelines for designing a digital twin for Li-ion battery: A reference methodology[J]. Energy, 2023, 284: 128699.
[2] Issa R, Badr MM, Shalash O, et al. A data-driven digital twin of electric vehicle Li-ion battery state-of-charge estimation enabled by driving behavior application programming interfaces[J]. Batteries, 2023, 9: 521.
[3] Schmitt J, Horstkötter I, Bäker B. State-of-health estimation by virtual experiments using recurrent decoder–encoder based lithium-ion digital battery twins trained on unstructured battery data[J]. Journal of Energy Storage, 2023, 58: 106335.
[4] Safavi V, Bazmohammadi N, Vasquez JC, et al. Battery state-of-health estimation: A step towards battery digital twins[J]. Electronics, 2024, 13: 587.
[5] Fu K, Hamacher T, Perić VS. Development of self-adaptive digital twin for battery monitoring and management system[J]. Electric Power Systems Research, 2024, 234: 110698.
[6] Lyu C, Xu S, Li J, et al. Digital twin modeling method for lithium-ion batteries based on data-mechanism fusion driving[J]. Green Energy and Intelligent Transportation, 2024, 3: 100162.
[7] Li W, Li Y, Garg A, et al. Enhancing real-time degradation prediction of lithium-ion battery: A digital twin framework with CNN-LSTM-attention model[J]. Energy, 2024, 286: 129681.
[8] Wang F, Zhai Z, Zhao Z, et al. Physics-informed neural network for lithium-ion battery degradation stable modeling and prognosis[J]. Nature Communications, 2024, 15: 4332.
[9] Ji C, Jin G, Zhang R. State of charge estimation for lithium-ion batteries based on a digital twin hybrid model[J]. Energy Reports, 2025, 13: 2174-2185.
[10] Salaorni D, Bianchi F, Colnago S, et al. A novel digital twin for battery energy storage systems in micro-grids[J]. Journal of Energy Storage, 2025, 132: 117745.
[11] Safavi V, Mohammadi Vaniar A, Bazmohammadi N, et al. A battery degradation-aware energy management system for agricultural microgrids[J]. Journal of Energy Storage, 2025, 108: 115059.
[12] Qasem M, Stubblefield J, Qandil M, et al. Real-time AI-enabled digital twin for battery health estimation and fast charging using partial-discharge data[J]. eTransportation, 2026, 27: 100528.
[13] Deng B, Zhong P, Jun K, et al. CHGNet as a pretrained universal neural network potential for charge-informed atomistic modelling[J]. Nature Machine Intelligence, 2023, 5: 1031-1041.
[14] Merchant A, Batzner S, Schoenholz SS, et al. Scaling deep learning for materials discovery[J]. Nature, 2023, 624: 80-85.
[15] Frank JT, Unke OT, Müller KR, et al. A Euclidean transformer for fast and stable machine learned force fields[J]. Nature Communications, 2024, 15: 6539.
[16] Zhang D, Bi H, Dai FZ, et al. Pretraining of attention-based deep learning potential model for molecular simulation[J]. npj Computational Materials, 2024, 10: 94.
[17] Batatia I, Benner P, Chiang Y, et al. A foundation model for atomistic materials chemistry[J]. Journal of Chemical Physics, 2025, 163: 184110.
[18] Radova M, Stark WG, Allen CS, et al. Fine-tuning foundation models of materials interatomic potentials with frozen transfer learning[J]. npj Computational Materials, 2025, 11: 237.
[19] Mazitov A, Bigi F, Kellner M, et al. PET-MAD as a lightweight universal interatomic potential for advanced materials modeling[J]. Nature Communications, 2025, 16: 10653.
[20] Loew A, Sun D, Wang HC, et al. Universal machine learning interatomic potentials are ready for phonons[J]. npj Computational Materials, 2025, 11: 178.
[21] Kabylda A, Frank JT, Suárez-Dou S, et al. Molecular simulations with a pretrained neural network and universal pairwise force fields[J]. Journal of the American Chemical Society, 2025, 147: 33723-33734.
[22] Wang R, Guo M, Gao Y, et al. A pre-trained deep potential model for sulfide solid electrolytes with broad coverage and high accuracy[J]. npj Computational Materials, 2025, 11: 266.
[23] Lai G, Zhang R, Fang C, et al. Machine-learning-accelerated mechanistic exploration of interface modification in lithium metal anode[J]. npj Computational Materials, 2025, 11: 245.
[24] Chen J, Fang M, Wu Q, et al. Insights into the atomic mechanism of lithium-ion diffusion in Li6PS5Cl via a machine learning potential[J]. Chemistry of Materials, 2025, 37: 591-599.
[25] Ma P, Kumar R, Wang KH, et al. Active learning accelerates electrolyte solvent screening for anode-free lithium metal batteries[J]. Nature Communications, 2025, 16: 8396.
[26] van der Oord C, Sachs M, Kovács DP, et al. Hyperactive learning for data-driven interatomic potentials[J]. npj Computational Materials, 2023, 9: 168.
[27] Liu K, Wei Z, Gao W, et al. Heterogeneous ensemble enables a universal uncertainty metric for atomistic foundation models[J]. npj Computational Materials, 2026, 12: 34.
[28] Kim K, Adelstein N, Dive A, et al. Probing degradation at solid-state battery interfaces using machine-learning interatomic potential[J]. Energy Storage Materials, 2024, 73: 103842.
[29] Xu G, Jiang M, Li J, et al. Machine learning-accelerated discovery and design of electrode materials and electrolytes for lithium ion batteries[J]. Energy Storage Materials, 2024, 72: 103710.
[30] Mortazavi B. Recent advances in machine learning-assisted multiscale design of energy materials[J]. Advanced Energy Materials, 2025, 15: 2403876.
[31] Poul M, Huber L, Neugebauer J. Automated generation of structure datasets for machine learning potentials and alloys[J]. npj Computational Materials, 2025, 11: 174.
[32] Kulichenko M, Barros K, Lubbers N, et al. Uncertainty-driven dynamics for active learning of interatomic potentials[J]. Nature Computational Science, 2023, 3: 230-239.
[33] Perez D, Subramanyam A P A, Swinburne T D. Uncertainty quantification for misspecified machine learned interatomic potentials[J]. npj Computational Materials, 2025, 11: 263.
[34] Qi J, Ko T W, Wood B C, et al. Robust training of machine learning interatomic potentials with dimensionality reduction and stratified sampling[J]. npj Computational Materials, 2024, 10: 43.
[35] Zaverkin V, Holzmüller D, Christiansen H, et al. Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials[J]. npj Computational Materials, 2024, 10: 83.
[36] Subramanyam A P A, Perez D. Information-entropy-driven generation of material-agnostic datasets for machine-learning interatomic potentials[J]. npj Computational Materials, 2025, 11: 218.
[37] Shaidu Y, Pellegrini F, Küçükbenli E, et al. Incorporating long-range electrostatics in neural network potentials via variational charge equilibration from shortsighted ingredients[J]. npj Computational Materials, 2024, 10: 47.
[38] Cheng B. Latent Ewald summation for machine learning of long-range interactions[J]. npj Computational Materials, 2025, 11: 80.
[39] Bistri D, Di Leo C V. A continuum electro-chemo-mechanical gradient theory coupled with damage: Application to Li-metal filament growth in all-solid-state batteries[J]. Journal of the Mechanics and Physics of Solids, 2023, 174: 105252.
[40] Taghikhani K, Huber W, Weddle P J, et al. Modeling coupled electro-chemo-mechanical phenomena within all-solid-state battery composite cathodes[J]. Journal of the Mechanics and Physics of Solids, 2025, 198: 106060.
[41] Chen W X, Allen J M, Rezaei S, et al. Cohesive phase-field chemo-mechanical simulations of inter- and trans-granular fractures in polycrystalline NMC cathodes via image-based 3D reconstruction[J]. Journal of Power Sources, 2024, 596: 234054.
[42] Gao X, Zhang Y, Liew K M. Electro-chemo-thermo-mechanical phase-field model for lithium penetration in solid electrolytes[J]. International Journal of Mechanical Sciences, 2025, 307: 110913.
[43] Zeng Q, Rezaei S, Carrillo L, et al. Chemomechanical damage prediction from phase-field simulation video sequences using a deep-learning-based methodology[J]. iScience, 2024, 27(9): 110822.
[44] Wang X, Li J, Wang R, et al. Lithium battery digital twin model with incremental learning[J]. CHAIN, 2024, 1(3): 249-262.
[45] Zhao K, Liu Y, Zhou Y, et al. Digital twin-supported battery state estimation based on TCN-LSTM neural networks and transfer learning[J]. CSEE Journal of Power and Energy Systems, 2025, 11(2): 567-579.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Yan Li, Duochang Tang
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This is an open-access article distributed under the terms of the Creative Commons Attribution CC BY 4.0 license, which permits unlimited use, distribution, and reproduction in any medium so long as the original work is properly cited.
