pith. sign in

arxiv: 2605.29560 · v1 · pith:FO3BCQXPnew · submitted 2026-05-28 · 💻 cs.AI

Battery-Sim-Agent: Leveraging LLM-Agent for Inverse Battery Parameter Estimation

Jiawei Chen , Xiaofan Gui , Shikai Fang , Shengyu Tao , Shun Zheng , Weiqing Liu , Jiang Bian This is my paper

Pith reviewed 2026-06-29 07:57 UTC · model grok-4.3

classification 💻 cs.AI
keywords battery parameter estimationLLM agentinverse problemdigital twinblack-box optimizationscientific reasoningdegradation modeling
0
0 comments X

The pith

An LLM agent estimates battery parameters more accurately than Bayesian optimization by reasoning over simulator feedback.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper reframes battery parameter estimation, needed for high-fidelity digital twins, as a reasoning task instead of a sample-inefficient black-box optimization problem. It introduces an LLM agent that closes the loop with a simulator: the agent reads multi-modal outputs, forms hypotheses about mismatches, and proposes parameter changes in a scientist-like workflow. On benchmarks covering multiple chemistries, conditions, and difficulty levels, this agent identifies parameters more accurately than strong baselines. The same loop handles long-horizon degradation fitting and real experimental data without domain-specific fine-tuning.

Core claim

Battery-Sim-Agent reframes the inverse parameter estimation task as closed-loop reasoning: an LLM agent receives rich simulator feedback, constructs physically grounded hypotheses to explain observed discrepancies, and issues structured parameter updates that progressively reduce error, outperforming Bayesian optimization and related BBO methods across a constructed benchmark suite of varied battery chemistries and operating regimes.

What carries the argument

The LLM agent that interprets multi-modal simulator feedback and generates hypothesis-driven parameter updates in a closed loop.

If this is right

  • Accurate parameters obtained this way produce digital twins that better match real battery behavior under diverse conditions.
  • The same reasoning loop extends to long-horizon degradation modeling tasks that traditional optimizers handle poorly.
  • The framework works directly on real-world battery measurement datasets without requiring chemistry-specific retraining.
  • Replacing blind search with hypothesis generation reduces the number of simulator calls needed to reach usable accuracy.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Similar agent loops could be tested on other inverse problems that combine expensive simulators with partial physical knowledge.
  • The approach suggests a route to hybrid systems where an LLM proposes candidate updates that a conventional optimizer then refines.
  • If the reasoning step generalizes, it could lower the barrier for non-experts to calibrate complex physics models.

Load-bearing premise

The LLM already holds enough pre-trained knowledge to turn simulator outputs into reliable physical hypotheses and parameter suggestions without hallucination or extra training.

What would settle it

Run the agent on a simulator with known ground-truth parameters and measure whether final estimated values converge to those known values within a stated tolerance after a fixed number of iterations.

Figures

Figures reproduced from arXiv: 2605.29560 by Jiang Bian, Jiawei Chen, Shengyu Tao, Shikai Fang, Shun Zheng, Weiqing Liu, Xiaofan Gui.

Figure 1
Figure 1. Figure 1: The closed-loop workflow of Battery-Sim-Agent. The agent proposes parameters for the PyBaMM simulator. The simulator’s output is then compared against target data to generate structured, multi-modal feedback (Sec. 3.2), which the agent analyzes using its dynamic memory (Sec. 3.3) to reason about the next parameter update. 3.3 Dynamic Memory with Knowledge Warm-up The agent’s ability to reason effectively r… view at source ↗
Figure 2
Figure 2. Figure 2: Main results on first-cycle calibration. Our [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Performance across C-rates. Comparison of different methods across various charge/discharge protocols. Each subplot [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Convergence analysis. Evolution of error metrics over optimization iterations for GPT-O3 on degradation fitting (left) [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Comparison of Warm-up Strategies. The boxplots il [PITH_FULL_IMAGE:figures/full_fig_p014_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Current–time and voltage–time curves for Experi [PITH_FULL_IMAGE:figures/full_fig_p015_7.png] view at source ↗
Figure 6
Figure 6. Figure 6: Typical Convergence Behaviors on Real-World Data. [PITH_FULL_IMAGE:figures/full_fig_p015_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: Best-so-far RMSE and MAPE over iterations for [PITH_FULL_IMAGE:figures/full_fig_p016_8.png] view at source ↗
read the original abstract

Parameterizing high-fidelity "digital twins" of batteries is a critical yet challenging inverse problem that hinders the pace of battery innovation. Prevailing methods formulate this as a black-box optimization (BBO) task, employing algorithms that are sample-inefficient and blind to the underlying physics. In this work, we introduce a new paradigm that reframes the inverse problem as a reasoning task, and present Battery-Sim-Agent, the first framework to deploy a Large Language Model (LLM) agent in a closed loop with a high-fidelity battery simulator. The agent mimics a human scientist's workflow: it interprets rich, multi-modal feedback from the simulator, forms physically-grounded hypotheses to explain discrepancies, and proposes structured parameter updates. On a systematically constructed benchmark suite spanning diverse battery chemistries, operating conditions, and difficulty levels, our agent significantly outperforms strong BBO baselines like Bayesian optimization in identifying accurate parameters. We further demonstrate the framework's capability in complex long-horizon degradation fitting tasks and validate its practical applicability on real-world battery datasets. Our results highlight the promise of LLM-agents as reasoning-based optimizers for scientific discovery and battery parameter estimation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 0 minor

Summary. The manuscript introduces Battery-Sim-Agent, an LLM-agent framework that operates in closed loop with a high-fidelity battery simulator to solve the inverse parameter estimation problem for battery digital twins. The agent interprets multi-modal simulator feedback, generates physically-grounded hypotheses, and proposes structured parameter updates. The central claim is that this reasoning-based approach significantly outperforms strong black-box optimization baselines such as Bayesian optimization on a systematically constructed benchmark spanning diverse chemistries, conditions, and difficulty levels, while also handling long-horizon degradation fitting and real-world datasets.

Significance. If the performance claims are substantiated with quantitative evidence, the work could mark a meaningful shift from sample-inefficient BBO methods toward agentic, physics-informed reasoning for scientific inverse problems. The conceptual reframing and use of simulator feedback are potentially valuable contributions to both battery modeling and LLM-agent applications in domain-specific optimization.

major comments (3)
  1. [Abstract] Abstract: the claim that the agent 'significantly outperforms strong BBO baselines like Bayesian optimization' is presented without any quantitative metrics, error values, success rates, or tabulated comparisons, rendering the central empirical claim unverifiable from the provided text.
  2. [Abstract] The manuscript supplies no methodological details on benchmark construction, parameter spaces, success metrics, controls for LLM stochasticity, or exclusion criteria, which are load-bearing for assessing whether the reported outperformance is robust.
  3. [Abstract] No discussion or ablation addresses the weakest assumption that the base LLM can reliably interpret multi-modal feedback and generate physically-grounded updates without hallucination or domain-specific fine-tuning.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive feedback. We address each major comment below, clarifying where details appear in the manuscript and indicating revisions to strengthen the abstract and related sections.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that the agent 'significantly outperforms strong BBO baselines like Bayesian optimization' is presented without any quantitative metrics, error values, success rates, or tabulated comparisons, rendering the central empirical claim unverifiable from the provided text.

    Authors: We agree that the abstract would be strengthened by including key quantitative indicators. The full manuscript reports these metrics (including mean parameter error, success rates, and direct comparisons to Bayesian optimization) in the Experiments section and associated tables. We will revise the abstract to incorporate representative quantitative results while respecting length constraints. revision: yes

  2. Referee: [Abstract] The manuscript supplies no methodological details on benchmark construction, parameter spaces, success metrics, controls for LLM stochasticity, or exclusion criteria, which are load-bearing for assessing whether the reported outperformance is robust.

    Authors: Detailed descriptions of benchmark construction, parameter spaces, success metrics, and controls for stochasticity (multiple independent runs with varied seeds) are provided in the Methods and Experimental Setup sections, with additional controls noted in the supplementary material. To improve self-containment of the abstract, we will add a concise summary of the benchmark suite and primary success criteria. revision: partial

  3. Referee: [Abstract] No discussion or ablation addresses the weakest assumption that the base LLM can reliably interpret multi-modal feedback and generate physically-grounded updates without hallucination or domain-specific fine-tuning.

    Authors: This observation is correct; the manuscript does not contain an explicit ablation on LLM reliability or hallucination rates. We will add a dedicated paragraph in the Discussion and Limitations sections addressing this assumption, including observed failure modes from the experiments and plans for future verification steps or lightweight fine-tuning. revision: yes

Circularity Check

0 steps flagged

No significant circularity; empirical framework with no derivations or self-referential reductions

full rationale

The paper introduces an LLM-agent framework for inverse battery parameter estimation and claims empirical outperformance versus BBO baselines on a constructed benchmark. No equations, derivations, or fitted parameters are presented in the provided text. The central claim rests on external comparison to independent baselines rather than any self-definition, self-citation chain, or renaming of known results. The benchmark and success metrics are described as systematically constructed but are not shown to reduce to quantities defined by the method itself. This is a standard non-circular empirical contribution.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review based on abstract only; the central claim rests on the unverified assumption that the LLM can perform physics-grounded reasoning from simulator outputs.

axioms (1)
  • domain assumption LLM can interpret rich, multi-modal feedback from the simulator, form physically-grounded hypotheses to explain discrepancies, and propose structured parameter updates
    This is the core workflow described in the abstract as mimicking a human scientist.

pith-pipeline@v0.9.1-grok · 5745 in / 1251 out tokens · 32933 ms · 2026-06-29T07:57:49.255707+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

41 extracted references · 20 canonical work pages · 1 internal anchor

  1. [1]

    iMOE: Prediction of Second-Life Battery Degradation Trajectory Using Interpretable Mixture of Experts.Nature Communications(2026)

    2026. iMOE: Prediction of Second-Life Battery Degradation Trajectory Using Interpretable Mixture of Experts.Nature Communications(2026). https://www. nature.com/articles/s41467-026-69369-1

    2026

  2. [2]

    Attia, Eric Moch, and Patrick K

    Peter M. Attia, Eric Moch, and Patrick K. Herring. 2025. Challenges and Op- portunities for High-Quality Battery Production at Scale. 16, 1 (2025), 611. doi:10.1038/s41467-025-55861-7

    work page doi:10.1038/s41467-025-55861-7 2025

  3. [3]

    Balog and Ali Davoudi

    Robert S. Balog and Ali Davoudi. 2013. Batteries, Battery Management , and Battery Charging Technology. InTransportation Technologies for Sustainability. Springer, New York, NY, 122–157. doi:10.1007/978-1-4614-5844-9_822

    work page doi:10.1007/978-1-4614-5844-9_822 2013

  4. [4]

    S Blaifi, S Moulahoum, I Colak, and W Merrouche. 2016. An enhanced dynamic model of battery using genetic algorithm suitable for photovoltaic applications. Applied Energy169 (2016), 888–898

    2016

  5. [5]

    Dhammika Widanage, and Emma Kendrick

    Chang-Hui Chen, Ferran Brosa Planella, Kieran O’Regan, Dominika Gastol, W. Dhammika Widanage, and Emma Kendrick. 2020. Development of Ex- perimental Techniques for Parameterization of Multi-scale Lithium-ion Bat- tery Models.Journal of The Electrochemical Society167, 8 (may 2020), 080534. doi:10.1149/1945-7111/ab9050

    work page doi:10.1149/1945-7111/ab9050 2020

  6. [6]

    Marc Doyle, Thomas F Fuller, and John Newman. 1993. Modeling of galvanos- tatic charge and discharge of the lithium/polymer/insertion cell.Journal of the Electrochemical society140, 6 (1993), 1526

    1993

  7. [7]

    Madeleine Ecker, Stefan Käbitz, Izaro Laresgoiti, and Dirk Uwe Sauer. 2015. Parameterization of a Physico-Chemical Model of a Lithium-Ion Battery: II. Model Validation.Journal of The Electrochemical Society162, 9 (June 2015), A1849. doi:10.1149/2.0541509jes

    work page doi:10.1149/2.0541509jes 2015

  8. [8]

    Madeleine Ecker, Thi Kim Dung Tran, Philipp Dechent, Stefan Käbitz, Alexander Warnecke, and Dirk Uwe Sauer. 2015. Parameterization of a Physico-Chemical Model of a Lithium-Ion Battery: I. Determination of Parameters.Journal of The Electrochemical Society162, 9 (June 2015), A1836. doi:10.1149/2.0551509jes

    work page doi:10.1149/2.0551509jes 2015

  9. [9]

    Gopinath, S

    R. Gopinath, S. Santhanagopalan, and Richard D. Braatz. 2016. An Inverse Method for Estimating the Electrochemical Parameters of Lithium-Ion Batteries.Journal of The Electrochemical Society163, 14 (2016), A3045–A3054

    2016

  10. [10]

    Ahmad Hamdan, Cosmas Daudu, Adefunke Fabuyide, Emmanuel Etukudoh, and Sedat Sonko. 2024. Next-Generation Batteries and U.S. Energy Storage: A Comprehensive Review: Scrutinizing Advancements in Battery Technology, Their Role in Renewable Energy, and Grid Stability. 21 (2024), 1984–1998. doi:10. 30574/wjarr.2024.21.1.0256

    2024

  11. [11]

    Nikolaus Hansen, Youhei Akimoto, and Petr Baudis. 2019. CMA-ES/pycma on Github. Zenodo, DOI:10.5281/zenodo.2559634. doi:10.5281/zenodo.2559634

    work page doi:10.5281/zenodo.2559634 2019

  12. [12]

    Wei He, Nicholas Williard, Michael Osterman, and Michael Pecht. 2011. Prognos- tics of Lithium-Ion Batteries Based on Dempster–Shafer Theory and the Bayesian Monte Carlo Method.Journal of Power Sources196, 23 (Dec. 2011), 10314–10321. doi:10.1016/j.jpowsour.2011.08.040

    work page doi:10.1016/j.jpowsour.2011.08.040 2011

  13. [13]

    Ming Hu, Chenglong Ma, Wei Li, Wanghan Xu, Jiamin Wu, Jucheng Hu, Tianbin Li, Guohang Zhuang, Jiaqi Liu, Yingzhou Lu, Ying Chen, Chaoyang Zhang, Cheng Tan, Jie Ying, Guocheng Wu, Shujian Gao, Pengcheng Chen, Jiashi Lin, Haitao Wu, Lulu Chen, Fengxiang Wang, Yuanyuan Zhang, Xiangyu Zhao, Feilong Tang, Encheng Su, Junzhi Ning, Xinyao Liu, Ye Du, Changkai Ji...

    work page arXiv 2025

  14. [14]

    Benben Jiang, Marc D Berliner, Kun Lai, Patrick A Asinger, Hongbo Zhao, Patrick K Herring, Martin Z Bazant, and Richard D Braatz. 2022. Fast charging design for Lithium-ion batteries via Bayesian optimization.Applied Energy307 (2022), 118244

    2022

  15. [15]

    Siyi Liu, Chen Gao, and Yong Li. 2024. Large language model agent for hyper- parameter optimization.arXiv preprint arXiv:2402.01881(2024)

    work page arXiv 2024

  16. [16]

    Dirk Magnor and Dirk Uwe Sauer. 2016. Optimization of PV battery systems using genetic algorithms.Energy Procedia99 (2016), 332–340

    2016

  17. [17]

    Marquis, Valentin Sulzer, Robert Timms, Colin P

    Scott G. Marquis, Valentin Sulzer, Robert Timms, Colin P. Please, and S. Jon Chapman. 2019. An Asymptotic Derivation of a Single Particle Model with Electrolyte.Journal of The Electrochemical Society166, 15 (Nov. 2019), A3693. doi:10.1149/2.0341915jes

    work page doi:10.1149/2.0341915jes 2019

  18. [18]

    2024.SimLM: Can Language Models Infer Parameters of Physical Systems?arXiv:2312.14215 [cs] doi:10.48550/ arXiv.2312.14215

    Sean Memery, Mirella Lapata, and Kartic Subr. 2024.SimLM: Can Language Models Infer Parameters of Physical Systems?arXiv:2312.14215 [cs] doi:10.48550/ arXiv.2312.14215

    work page arXiv 2024

  19. [19]

    Bo Ni and Markus J Buehler. 2024. MechAgents: Large language model multi- agent collaborations can solve mechanics problems, generate new data, and integrate knowledge.Extreme Mechanics Letters67 (2024), 102131

    2024

  20. [20]

    Miles Olson, Elizabeth Santorella, Louis C. Tiao, Sait Cakmak, David Eriksson, Mia Garrard, Sam Daulton, Maximilian Balandat, Eytan Bakshy, Elena Kashtelyan, Zhiyuan Jerry Lin, Sebastian Ament, Bernard Beckerman, Eric Onofrey, Paschal Igusti, Cristian Lara, Benjamin Letham, Cesar Cardoso, Shiyun Sunny Shen, Andy Chenyuan Lin, and Matthew Grange. 2025. Ax:...

    2025

  21. [21]

    OpenAI. 2025. OpenAI o3 and o4-mini System Card. https://cdn.openai.com/ pdf/2221c875-02dc-4789-800b-e7758f3722c1/o3-and-o4-mini-system-card.pdf

    2025

  22. [22]

    OpenAI, Sandhini Agarwal, Lama Ahmad, Jason Ai, Sam Altman, Andy Apple- baum, Edwin Arbus, Rahul K. Arora, Yu Bai, Bowen Baker, Haiming Bao, Boaz Barak, Ally Bennett, Tyler Bertao, Nivedita Brett, Eugene Brevdo, Greg Brockman, Sebastien Bubeck, Che Chang, Kai Chen, Mark Chen, Enoch Cheung, Aidan Clark, Dan Cook, Marat Dukhan, Casey Dvorak, Kevin Fives, Vl...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2508.10925 2025

  23. [24]

    Wendy Pantoja, Jaime Andres Perez-Taborda, and Alba Avila. 2022. Tug-of-War in the Selection of Materials for Battery Technologies.Batteries8, 9 (Sept. 2022),

    2022

  24. [25]

    doi:10.3390/batteries8090105

    work page doi:10.3390/batteries8090105

  25. [26]

    Prada, D

    E. Prada, D. Di Domenico, Y. Creff, J. Bernard, V. Sauvant-Moynot, and F. Huet

  26. [27]

    2013), A616

    A Simplified Electrochemical and Thermal Aging Model of LiFePO4- Graphite Li-ion Batteries: Power and Capacity Fade Simulations.Journal of The Electrochemical Society160, 4 (Feb. 2013), A616. doi:10.1149/2.053304jes

    work page doi:10.1149/2.053304jes 2013

  27. [28]

    Prasad, A

    K. Prasad, A. Rahimian, and M. Fowler. 2015. Inverse parameter determination in the development of an optimized lithium iron phosphate–Graphite battery discharge model.Journal of Power Sources273 (2015), 1348–1359

    2015

  28. [30]

    Ana-Irina Stroe, Daniel-Loan Stroe, Vaclav Knap, Maciej Swierczynski, and Remus Teodorescu. 2018. Accelerated Lifetime Testing of High Power Lithium Titanate Oxide Batteries. In2018 IEEE Energy Conversion Congress and Exposition (ECCE). 3857–3863. doi:10.1109/ECCE.2018.8557416

    work page doi:10.1109/ecce.2018.8557416 2018

  29. [31]

    Subramanian and Richard D

    Venkat R. Subramanian and Richard D. Braatz. 2013. Modeling and Simulation of Lithium-Ion Batteries from a Systems Engineering Perspective.Journal of The Electrochemical Society160, 4 (2013), R93–R108

    2013

  30. [32]

    Marquis, Robert Timms, Martin Robinson, and S

    Valentin Sulzer, Scott G. Marquis, Robert Timms, Martin Robinson, and S. Jon Chapman. 2021. PyBaMM: Python Battery Mathematical Modelling.Journal of Open Research Software9, 1 (2021), 14

    2021

  31. [33]

    Zechang Sun, Yuan-Sen Ting, Yaobo Liang, Nan Duan, Song Huang, and Zheng Cai. 2024. Interpreting multi-band galaxy observations with large language model-based agents.arXiv preprint arXiv:2409.14807(2024)

    work page arXiv 2024

  32. [34]

    Xizhe Wang and Benben Jiang. 2023. Multi-objective optimization for fast charg- ing design of lithium-ion batteries using constrained Bayesian optimization. Journal of Power Sources584 (2023), 233602

    2023

  33. [35]

    Jiaqi Wei, Yuejin Yang, Xiang Zhang, Yuhan Chen, Xiang Zhuang, Zhangyang Gao, Dongzhan Zhou, Guangshuai Wang, Zhiqiang Gao, Juntai Cao, et al. 2025. From AI for Science to Agentic Science: A Survey on Autonomous Scientific Discovery.arXiv preprint arXiv:2508.14111(2025)

    work page arXiv 2025

  34. [36]

    Mengsong Wu, YaFei Wang, Yidong Ming, Yuqi An, Yuwei Wan, Wenliang Chen, Binbin Lin, Yuqiang Li, Tong Xie, and Dongzhan Zhou. 2025. ChemAgent: Enhancing LLMs for Chemistry and Materials Science through Tree-Search Based Tool Learning.arXiv preprint arXiv:2506.07551(2025). Battery-Sim-Agent: Leveraging LLM-Agent for Inverse Battery Parameter Estimation KDD...

    work page arXiv 2025

  35. [37]

    Yinjiao Xing, Eden W. M. Ma, Kwok-Leung Tsui, and Michael Pecht. 2013. An Ensemble Model for Predicting the Remaining Useful Performance of Lithium-Ion Batteries.Microelectronics Reliability53, 6 (June 2013), 811–820. doi:10.1016/j. microrel.2012.12.003

    work page doi:10.1016/j 2013

  36. [38]

    Jones, and Michael A

    Wenjie Xu, Masaki Adachi, Colin N. Jones, and Michael A. Osborne. 2024.Principled Bayesian Optimisation in Collaboration with Human Experts. arXiv:2410.10452 [cs] doi:10.48550/arXiv.2410.10452

    work page doi:10.48550/arxiv.2410.10452 2024

  37. [39]

    Han Zhang, Xiaofan Gui, Shun Zheng, Ziheng Lu, Yuqi Li, and Jiang Bian. 2024. BatteryML: An Open-source Platform for Machine Learning on Battery Degrada- tion. InThe Twelfth International Conference on Learning Representations

    2024

  38. [40]

    Liqiang Zhang, Lixin Wang, Gareth Hinds, Chao Lyu, Jun Zheng, and Junfu Li

  39. [41]

    Multi-objective optimization of lithium-ion battery model using genetic algorithm approach.Journal of Power Sources270 (2014), 367–378

    2014

  40. [42]

    ground truth

    Wenhua Zuo, Huihuo Zheng, Tanjin He, Venkatram Vishwanath, Maria KY Chan, Rick L Stevens, Khalil Amine, and Gui-Liang Xu. 2025. Large language models for batteries.Joule9, 8 (2025). A Reproducibility statement We have taken several measures to ensure the reproducibility of our results. All experiments were conducted with fixed random seeds, and key experi...

    2025

  41. [43]

    • A higher SEI solvent diffusivity (SEI_solvent_diffusivity_m2_s-1) increases the degradation rate and magnitude

    generally results in larger initial capacity and impedance decay, with a downward-convex curve. • A higher SEI solvent diffusivity (SEI_solvent_diffusivity_m2_s-1) increases the degradation rate and magnitude. • A higher EC diffusivity (EC_diffusivity_m2_s-1) accelerates the degradation rate and results in a downward-convex curve. • A higher initial SEI t...

    2007