Pith. sign in

REVIEW 3 major objections 5 minor 57 references

A Primer on Bayesian Parameter Estimation and Model Selection for Battery Simulators

T0 review · 3 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Two Bayesian algorithms, SOBER and BASQ, can parameterize and rank battery models from test data in a fraction of the simulator runs that Markov-chain sampling requires, and the paper demonstrates the pair on six battery problems.

desk verdict Useful Bayesian primer for battery modelers, but Example 6's model-ranking claims rest on an arbitrary evidence rescaling the authors themselves flag; the 'automatically, accurately sort models' conclusion is not yet supported. read the letter →

arxiv 2512.10055 v1 pith:ZMF4XY75 submitted 2025-12-10 stat.ME physics.data-anstat.AP

Yannick Kuhn , Masaki Adachi , Micha Philipp , David A. Howey , Birger Horstmann This is my paper · ORCID
classification stat.MEphysics.data-anstat.AP MSC 62F1562L0562P30
keywords BayesianinferenceModelselectionBatterysimulationLikelihood-freeoptimisationquadratureGaussianprocessesElectrochemicalimpedancespectroscopy
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that combining two Bayesian algorithms—SOBER for parameter estimation and BASQ for model evidence—turns battery model comparison into an automated, uncertainty-quantified procedure. The central claim is that this pairing works with only a few hundred simulator evaluations, making Bayesian inference practical for expensive physics-based battery simulators. The authors demonstrate the approach on battery sizing, voltage relaxation analysis, knee-point detection in degradation data, inverse surrogate training, electrolyte identifiability, and impedance model selection. If the claim holds, battery researchers can replace slow MCMC sampling and heuristic curve fitting with posterior distributions and evidence-based model ranking, which matters because new battery chemistries generate a growing range of candidate models.

What carries the argument

SOBER is the parameter-estimation engine: it models the discrepancy between simulator output and data with a Gaussian process, uses an uncertainty-sampling acquisition function to choose batches of simulator runs, and returns an approximate posterior distribution over parameters. BASQ is the model-selection engine: it computes the marginal-likelihood integral (evidence) over that posterior, producing a scalar score for comparing candidate models. The pairing works because SOBER makes inference sample-efficient enough for expensive battery simulators, while BASQ turns each parameterisation into a model-comparison criterion that also serves as a convergence check.

What would settle it

Run a battery test case through SOBER+BASQ under the paper's fixed evaluation budgets and compare the resulting posterior and evidence ranking against a gold-standard MCMC posterior and a precise evidence estimate (e.g., thermodynamic integration or annealed importance sampling). Disagreement beyond the stated credibility intervals on a dataset where the candidate models are known to be measurably different would directly contradict the central claim.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that a likelihood-free Bayesian inference pipeline—SOBER, which frames parameter fitting as Bayesian optimisation with an uncertainty-sampling acquisition function, and BASQ, which evaluates the model-evidence integral—can be applied to battery simulators with modest numbers of model evaluations. The key demonstration is that model evidence, not just point fits, can be computed accurately enough to rank models: the one-knee vs. two-knee degradation model, the single-particle vs. full-transport impedance model, and the ordering of double-layer vs. SEI timescales are all decided by comparing evidence values. The same machinery also trains a global invers

Load-bearing premise

The preset number of simulator evaluations is large enough for SOBER and BASQ to converge to the true posterior and to evidence values accurate enough for ranking; this is mostly assumed rather than verified per example, and some reported evidence intervals are wide enough that rankings by means may not be trustworthy.

Editorial extensions

If this is right

  • Battery researchers can rank candidate physics-based models by computed evidence, replacing manual inspection of fit curves or single-point RMSE comparisons.
  • Parameterisation outputs include posterior distributions and correlation matrices, so identifiability problems—parameters that cannot be separated from data—become visible directly from the fit.
  • The same evidence calculation can flag when data are too weak to support a model choice, since wide evidence intervals signal that ranking by mean values is unreliable.
  • Active learning with SOBER can build inverse surrogates that map measured signals back to operating conditions or material parameters with far fewer simulator runs than naive design-of-experiments.
  • The approach generalises naturally to automated laboratories: the algorithm can decide which experiment to run next based on where the surrogate or evidence is most uncertain.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The evidence-based workflow could serve as a screening tool in high-throughput materials discovery, where the evidence value itself is a compact, uncertainty-aware descriptor of whether an experiment carries information about a proposed model.
  • A testable extension would apply the same pipeline to streaming degradation data, checking whether the evidence for a second knee point rises before the knee is visually apparent, enabling early warning in second-life battery applications.
  • The fixed evaluation budgets used throughout the paper suggest that the method may be less robust when the posterior is multimodal, a case the paper only briefly mentions; applying the approach to deliberately multimodal synthetic problems would test that boundary.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper is a tutorial-style application paper introducing the Bayesian optimisation algorithm SOBER and the Bayesian quadrature algorithm BASQ to the battery modelling community, with the software wrapped into the PyBOP package. A brief derivation of Bayesian inference and likelihood-free inference is given, followed by six worked examples: battery sizing, voltage relaxation parameterisation, knee-point model selection from degradation data, inverse modelling via a global inverse surrogate, electrolyte transport identifiability analysis, and impedance-based model selection between SPM and DFN models. The central claim is that SOBER and BASQ together allow users to 'automatically, accurately, and quickly sort through this plethora of models' (Conclusion), enabling uncertainty-quantified parameter estimation and evidence-based model comparison for battery simulators.

Significance. If the evidence calculations were fully valid, the paper would be a valuable methodological introduction for the battery community, with the practical contribution of open-source implementations and a wide range of demonstration cases. The first five examples are instructive and the predictive-posterior checks are a good practice. However, the key validation of the central model-selection claim in Example 6 is compromised: the DFN evidence is rescaled by an unexplained constant, and the reported evidence credible intervals are so wide that the model rankings are not robust. Because this example is the main evidence for the paper's headline claim, the manuscript needs substantial revision before the contribution can be considered established. The paper is more a primer and software showcase than a new theoretical contribution, but that is consistent with its stated purpose.

major comments (3)
  1. [Example 6, Table 6] The DFN evidence is multiplied by (sqrt(2*pi))^3 and reported as 'DFN (corr.)'. This factor is not part of the marginal likelihood p(D|M)=int p(D|theta,M)p(theta)dtheta that BASQ is designed to compute; that integral already contains any complexity/Occam penalty through the prior and the integrated likelihood. The multiplier is a constant independent of the data, the prior, and the posterior, so it cannot be derived merely from the normalisation of a Gaussian density. The raw mean-evidence ratios from Table 6(a) are approximately 3%, 5%, and 49% for the 17, 42, and 80 um electrodes; after multiplication they become 51%, 117%, and 792%. Thus the conclusion that the DFN is preferred for thicker electrodes is an artifact of this rescaling. The text itself is contradictory: it says one must 'correct' for the integral penalty by multiplying by (sqrt(2*pi))^3, yet a sentence later says 'We pre
  2. [Example 6, Result and Discussion, Table 6] The evidence credible intervals are extremely wide, often spanning more than an order of magnitude (e.g., SPM at 17 um: [0.6, 95]; DFN at 80 um: [0.3, 17]), and the intervals for competing models overlap substantially at 17 and 42 um. The statement that ranking by mean evidence is valid 'as long as we ensure the variances are smaller than the means' is not established; for several entries the interval width implies a variance comparable to or larger than the mean. No per-example convergence diagnostics are shown beyond two predictive-posterior plots for DFN fits. The fixed evaluation budget of 48 initial samples and 19 SOBER iterations at 48 samples each for all impedance examples, independent of the number of fitted parameters, is a strong assumption. The paper should demonstrate, at least for the model-selection examples, that the BASQ evidence estimates are converged with respect to i
  3. [Example 5, Method] The statement 'We fix kappa_e,peak and kappa_e,spread at 1.0, as we found them to be virtually unidentifiable in a first test run' is circular with respect to the example's stated aim of assessing whether electrolyte parameters are uniquely identifiable from a constant-current pulse. Excluding two of the four transport-shape parameters based on a preliminary run means the subsequent identifiability analysis covers only the remaining parameters and cannot make claims about the fixed ones. If the intended message is that these parameters are unidentifiable, the paper should demonstrate this from the full model and report the preliminary evidence. If they are fixed for numerical conditioning, the clustering and correlation results should be explicitly described as conditional on that choice, and the generality of the identifiability conclusion should be qualified.
minor comments (5)
  1. [Example 6, Table 6 caption] The caption does not define 'DFN (corr.)'. State explicitly that this column is the BASQ evidence multiplied by (sqrt(2*pi))^3, and distinguish it from the unmodified evidence in Table 6(a).
  2. [Example 6, text near Eq. (23)] The sentence 'We prefer comparing the determinant of Sigma between different models directly, without the arbitrary dimensionality penalty' conflicts with the rescaling actually applied. Rewrite to explain the exact relationship between the correction factor and the posterior covariance determinant, or remove the correction.
  3. [Example 4, Figure 7 and Eq. (21)] The Kronecker-structure multi-output GP notation in Eq. (21) is introduced without a definition of the kernel K or the fidelity kernel K_f. A reader unfamiliar with multi-fidelity GPs will not be able to reproduce this example.
  4. [Example 6, 'For all impedance examples...'] The fixed number of integration nodes for BASQ is '3 to the power of the number of fit parameters', which grows quickly (e.g., 729 nodes for three parameters, 531441 for six parameters). Please state whether this is the number of nodes per dimension or the total number, and how the evidence variance is estimated.
  5. [General] The paper uses the authors' own previously published algorithms (SOBER, ref. 12; BASQ, ref. 13; EP-BOLFI, ref. 4) and, in Example 6, the authors' own impedance data (ref. 53). This should be stated more explicitly in the main text so that readers understand that the demonstrations are not independent third-party validations.

Circularity Check

1 steps flagged · score 6.0 of 10

Example 6's model-selection conclusion is produced by a post hoc, data-independent sqrt(2π)^3 multiplier on all DFN evidences; without it the reported SPM/DFN rankings reverse.

  1. other [Section 'Example 6: Impedance model selection', Result and Discussion, text following Table 6]
    "The second way compares different models on the same data. To do so, we have to correct for the integral penalty for the extra 3 fit parameters for the DFN, multiplying the evidence by sqrt(2π)^3. ... We prefer comparing the determinant of Σ between different models directly, without the 'arbitrary' dimensionality penalty sqrt(2π)^(d2-d1)."

    Evidence is defined in Eq. (11) as p(D|M)=∫p(D|θ,M)p(θ)dθ, which already includes any prior-volume/parameter-dimension penalty. Multiplying only the DFN evidences by the constant sqrt(2π)^3≈15.75 is not derived from the BASQ evidence estimates, from the posterior covariance determinants, or from the data; it is a data-independent rescaling favoring the 6-parameter model. Raw mean evidence ratios from Table 6(a) are DFN/SPM ≈0.03, 0.05, 0.49 for 17, 42, 80 µm; after the multiplier they become ≈0.51, 1.17, 7.92, exactly the 'DFN (corr.)' percentages in Table 6(c). Thus the headline result that the DFN is preferred for thicker electrodes reduces to the chosen multiplier, not to the computed evidences.

full rationale

The paper is largely a self-contained primer and application showcase: SOBER and BASQ are self-cited, but they are used as off-the-shelf algorithms and the first five examples are independent demonstrations with external data or synthetic benchmarks. The one load-bearing circularity is in Example 6, the paper's real-data validation of the central claim that SOBER+BASQ automatically and accurately sort battery models. There, the authors explicitly alter the BASQ-computed DFN evidences by a constant sqrt(2π)^3, called a 'correction' for an 'integral penalty'. Since the evidence integral already contains the Occam/prior-volume penalty, this constant is not part of the evidence and is not derived from any posterior diagnostic. The raw Table 6(a) values show DFN/SPM ratios below 1 for all three electrode thicknesses, so without the multiplier the DFN is never preferred; with it, the DFN appears preferred for the two thicker electrodes. The model-selection conclusion therefore reduces to a post hoc rescaling rather than to the computed evidence. Other self-citations (refs 4, 12, 13, 53) are normal and do not by themselves carry the derivation, so the circularity is partial rather than total.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

No new physical entities are introduced. The free parameters listed are the two most load-bearing ad hoc choices; the axioms are the statistical and modeling assumptions that the pipeline and the example conclusions depend on.

free parameters (2)
  • DFN evidence rescaling factor = (√2π)^3 ≈ 15.75
    Arbitrary multiplier applied to DFN evidences in Table 6 to account for 3 extra parameters; not derived from the model evidence, and the paper calls the penalty 'arbitrary'.
  • κe,peak and κe,spread (electrolyte conductivity shape) = 1.0
    Fixed after a first test run in Example 5 because they were 'virtually unidentifiable'; a post hoc choice made on the same analysis.
assumptions (4)
  • domain assumption Gaussian process priors sufficiently model the discrepancy surface and the inverse mapping (Eq. 15, Eq. 21).
    The whole SOBER/BASQ pipeline assumes GP surrogates are good approximations for the battery simulator's RMSE and for (a,b)→(I,τ) and overpotential→electrolyte-parameter mappings; no validation of this smoothness in all examples.
  • domain assumption The pseudo-likelihood with adaptively decreasing ε approximates the true posterior within the fixed evaluation budget (Eq. 14-16).
    Theoretical convergence (refs 23,24) holds in the limit ε→0; the paper does not verify this limit was reached for each example.
  • ad hoc to paper Posterior distributions are approximately Gaussian for the covariance analysis and evidence rescaling.
    Invoked in Example 6: 'Identifiable optimisation tasks tend to a Gaussian posterior asymptotically per the central limit theorem', used to justify the (√2π)^d normalisation.
  • domain assumption The battery models (DFN, SPM, SPMe, SEI impedance) are structurally adequate for the datasets.
    Model selection compares only these candidate models; if all are misspecified, the evidence rankings may not identify the 'true' model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Primer on Bayesian Parameter Estimation and Model Selection for Battery Simulators." pith.science (2026). https://pith.science/paper/ZMF4XY75

@misc{pith2026251210055,
  author       = {Pith},
  title        = {Pith review of: A Primer on Bayesian Parameter Estimation and Model Selection for Battery Simulators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZMF4XY75}},
  note         = {Machine review of arXiv:2512.10055}
}
read the original abstract

Physics-based battery modelling has emerged to accelerate battery materials discovery and performance assessment. Its success, however, is still hindered by difficulties in aligning models to experimental data. Bayesian approaches are a valuable tool to overcome these challenges, since they enable prior assumptions and observations to be combined in a principled manner that improves numerical conditioning. Here we introduce two new algorithms to the battery community, SOBER and BASQ, that greatly speed up Bayesian inference for parameterisation and model comparison. We showcase how Bayesian model selection allows us to tackle data observability, model identifiability, and data-informed model development together. We propose this approach for the search for battery models of novel materials.

Figures

Figures reproduced from arXiv: 2512.10055 by the authors.

Figure 1
Figure 1. Illustration of the relation between model and [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Extra lifetime of a solar panel battery over [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 4
Figure 4. Correlation matrix of multivariate normal ap [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figures from the paper (9 more)
Figure 5
Figure 5. Figure 5: Measurement of silicon electrode voltage relax [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 7
Figure 7. Figure 7: Illustration for the way SOBER is utilised [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 6
Figure 6. Figure 6: Model selection example with (a) one-knee [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 9
Figure 9. Figure 9: Illustration for the way SOBER is utilised [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 11
Figure 11. Figure 11: Correlation matrix of one representative eval [PITH_FULL_IMAGE:figures/full_fig_p013_11.png]
Figure 10
Figure 10. Figure 10: Training data—model parameters for inverse [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]
Figure 12
Figure 12. Figure 12: Violin plots of active material diffusivities as [PITH_FULL_IMAGE:figures/full_fig_p015_12.png]
Figure 13
Figure 13. Figure 13: Predictive posterior of impedance parameter [PITH_FULL_IMAGE:figures/full_fig_p016_13.png]
Figure 14
Figure 14. Figure 14: Predictive posterior of impedance parame [PITH_FULL_IMAGE:figures/full_fig_p018_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 12 canonical work pages

  1. [1]

    Sethurajan, S

    A. Sethurajan, S. Krachkovskiy, G. Goward, B. Protas,Journal of Computational Chemistry2019, 40, 740–752, DOI 10.1002/jcc.25759

    work page doi:10.1002/jcc.25759

  2. [2]

    Aitio, S

    A. Aitio, S. G. Marquis, P. Ascencio, D. Howey, IFAC-PapersOnLine2020,53, 12497–12504, DOI 10.1016/j.ifacol.2020.12.1770

    work page doi:10.1016/j.ifacol.2020.12.1770 2020

  3. [3]

    J. M. Escalante, S. Sahu, J. M. Foster, B. Protas,The Electrochemical Society2021,168, 110519, DOI 10.1149/1945-7111/ac3159

    work page doi:10.1149/1945-7111/ac3159 1945

  4. [4]

    Y. Kuhn, H. Wolf, A. Latz, B. Horstmann,Batteries & Supercaps2023,6, e202200374, DOI 10.1002/ batt.202200374, https://doi.org/10.1002/batt. 202200374

    work page doi:10.1002/batt

  5. [5]

    Adachi, Y

    M. Adachi, Y. Kuhn, B. Horstmann, A. Latz, M. A. Osborne, D. A. Howey,IFAC-PapersOnLine2023, 56, 10521–10526, DOI 10.1016/j.ifacol.2023.10. 1073

    doi:10.1016/j.ifacol.2023.10 2023

  6. [6]

    X. Xia, E. Sivonxay, B. A. Helms, S. M. Blau, E. M. Chan,Nano Letters2023,23, 11129–11136, DOI 10.1021/acs.nanolett.3c03568

    work page doi:10.1021/acs.nanolett.3c03568

  7. [7]

    S. P. Stier, C. Kreisbeck, H. Ihssen, M. A. Popp, J. Hauch, K. Malek, M. Reynaud, T. P. Goumans, J. Carlsson, I. Todorov, L. Gold, A. R¨ader, W. Wen- zel, S. T. Bandesha, P. Jacques, F. Garcia-Moreno, O. Arcelus, P. Friederich, S. Clark, M. Maglione, A. Laukkanen, I. E. Castelli, J. Carrasco, M. C. Ca- banas, H. S. Stein, O. Ozcan, D. Elbert, K. Reuter, C...

  8. [8]

    Vogler, S

    M. Vogler, S. K. Steensen, F. F. Ram´ırez, L. Merker, J. Busk, J. M. Carlsson, L. H. Rieger, B. Zhang, F. Liot, G. Pizzi, F. Hanke, E. Flores, H. Hajiyani, S. Fuchs, A. Sanin, M. Gaber ˇsˇcek, I. E. Castelli, S. Clark, T. Vegge, A. Bhowmik, H. S. Stein,Ad- vanced Energy Materials2024,14, 2403263, DOI 10.1002/aenm.202403263, https://onlinelibrary. wiley.co...

    work page doi:10.1002/aenm.202403263

Show all 57 references
  1. [9]

    Sanin, J

    A. Sanin, J. K. Flowers, T. H. Piotrowiak, F. Felsen, L. Merker, A. Ludwig, D. Bresser, H. S. Stein,Ad- vanced Energy Materials2025, 2404961, DOI 10. 1002/aenm.202404961

  2. [10]

    Philipp, Y

    M. Philipp, Y. Kuhn, A. Latz, B. Horstmann, ChemSusChem2025, e202402336, DOI 10.1002/ cssc . 202402336, https : / / chemistry - europe . onlinelibrary . wiley . com / doi / 10 . 1002 / cssc . 202402336

  3. [11]

    Landesfeind, H

    J. Landesfeind, H. A. Gasteiger,Journal of The Electrochemical Society2019,166, A3079–A3097, DOI 10.1149/2.0571912jes

    doi

  4. [12]

    Adachi, S

    M. Adachi, S. Hayakawa, M. Jørgensen, S. Hamid, H. Oberhauser, M. A. Osborne,arXiv2024, http: //arxiv.org/abs/2404.12219

    arXiv

  5. [13]

    Adachi, S

    M. Adachi, S. Hayakawa, H. Oberhauser, M. Jørgensen, M. A. Osborne in Advances in Neu- ral Information Processing Systems 35 (NeurIPS 2022),2022, pp. 16533–16547, https://proceedings. neurips . cc / paper files / paper / 2022 / file / 697200c9d1710c2799720b660abd11bb - Paper -...

    2022

  6. [14]

    Aitio, D

    A. Aitio, D. A. Howey,Joule2021,5, 3204–3220, DOI 10.1016/j.joule.2021.11.006

    2021 doi

  7. [15]

    Dubarry, D

    M. Dubarry, D. Beck,Energies2021,14, 2371, DOI 10.3390/en14092371

    doi

  8. [16]

    Planden, N

    B. Planden, N. E. Courtier, M. Robinson, A. Khetarpal, F. B. Planella, D. A. Howey,arXiv2024, DOI 10.48550/arXiv.2412.15859, http://arxiv.org/ abs/2412.15859

    doi

  9. [17]

    A. M. Bizeray, J. H. Kim, S. R. Duncan, D. A. Howey,IEEE Transactions on Control Systems Technology2019,27, 1862–1877, DOI 10.1109/ TCST.2018.2838097

    2018

  10. [18]

    C. J. Geyer,Statistical Science1992,7, 473–511, https://www.jstor.org/stable/2246094

  11. [19]

    Diggle, R

    P. Diggle, R. Gratton,Journal of the Royal Statisti- cal Society Series B: Statistical Methodology1984, 46, 193–212

  12. [20]

    Rubin,The Annals of Statistics1984, 1151– 1172

    D. Rubin,The Annals of Statistics1984, 1151– 1172

  13. [21]

    Marin, P

    J.-M. Marin, P. Pudlo, C. Robert, R. Ryder,Statis- tics and computing2012,22, 1167–1180

  14. [22]

    Genton, E

    M. Genton, E. Ronchetti,Journal of the American Statistical Association2003,98, 67–76

  15. [23]

    Wilkinson,Statistical applications in genetics and molecular biology2013,12, 129–141

    R. Wilkinson,Statistical applications in genetics and molecular biology2013,12, 129–141

  16. [24]

    Frazier, G

    D. Frazier, G. Martin, C. Robert, J. Rousseau, Biometrika2018,105, 593–607

  17. [25]

    C. E. Rasmussen, C. K. I. Williams,Gaussian processes for machine learning, MIT Press,2006, p. 248

    2006

  18. [26]

    M. U. Gutmann, J. Corander,Journal of Machine Learning Research2016,17, 1–47, http://jmlr.org/ papers/v17/15-017.html

  19. [27]

    Garnett,Bayesian optimization, Cambridge University Press,2023

    R. Garnett,Bayesian optimization, Cambridge University Press,2023

    2023

  20. [28]

    Srinivas, A

    N. Srinivas, A. Krause, S. M. Kakade, M. W. Seeger, IEEE Transactions on Information Theory2012,58, 3250–3265, DOI 10.1109/TIT.2011.2182033

    2011

  21. [29]

    S. R. Chowdhury, A. Gopalan,Proceedings of the 34th International Conference on Machine Learn- ing2017,70, 844–853, https://proceedings.mlr. press/v70/chowdhury17a.html

  22. [30]

    T. P. Minka, A family of algorithms for approxi- mate Bayesian inference,2001, https://tminka. github.io/papers/ep/minka-thesis.pdf

    2001

  23. [31]

    J. M. Reniers, G. Mulder, D. A. Howey,Journal of The Electrochemical Society2019,166, A3189– A3200, DOI 10.1149/2.0281914jes

    doi

  24. [32]

    von Kolzenberg, A

    L. von Kolzenberg, A. Latz, B. Horstmann,Chem- SusChem2020,13, 3901, DOI 10 . 1002 / cssc . 202000867, http://arxiv.org/abs/2004.01458http: //dx.doi.org/10.1002/cssc.202000867

    2004 arXiv

  25. [33]

    Albertus, J

    P. Albertus, J. Newman,University of California Berkeley2007

  26. [34]

    Wycisk, G

    D. Wycisk, G. K. Mertin, M. Oldenburger, O. von Kessel, A. Latz,Journal of Energy Storage2024, 89, 111617, DOI 10.1016/j.est.2024.111617

    2024

  27. [35]

    S. D. Kang, W. C. Chueh,Journal of The Electro- chemical Society2021,168, 120504, DOI 10.1149/ 1945-7111/ac3940

    1945

  28. [36]

    K¨obbing, Y

    L. K¨obbing, Y. Kuhn, B. Horstmann,ACS Applied Materials & Interfaces2024,16, 67609–67619, DOI 10.1021/acsami.4c12976, http://arxiv.org/abs/ 2408.01106

    arXiv

  29. [37]

    H. Zhao, J. Meng, Q. Peng,Applied Energy2025, 381, 125214, DOI 10.1016/j.apenergy.2024.125214

    2024

  30. [38]

    S. Tao, M. Zhang, Z. Zhao, H. Li, R. Ma, Y. Che, X. Sun, L. Su, C. Sun, X. Chen, H. Chang, S. Zhou, Z. Li, H. Lin, Y. Liu, W. Yu, Z. Xu, H. Hao, S. Moura, X. Zhang, Y. Li, X. Hu, G. Zhou,Energy and Envi- ronmental Science2025,18, 1544–1559, DOI 10. 1039/d4ee03839h

  31. [39]

    J. Zhu, W. Weng, H. You, J. Zhang, Y. Wang, B. Jiang, C. Ji, X. Wei, H. Dai,Applied Energy2025, 401, DOI 10.1016/j.apenergy.2025.126692

    2025

  32. [40]

    P. M. Attia, A. Bills, F. B. Planella, P. Dechent, G. dos Reis, M. Dubarry, P. Gasper, R. Gilchrist, S. Greenbank, D. Howey, O. Liu, E. Khoo, Y. Preger, A. Soni, S. Sripad, A. G. Stefanopoulou, V. Sulzer, Journal of The Electrochemical Society2022,169, 060517, DOI 10.1149/1945...

    1945 doi

  33. [41]

    Y. Ni, X. Li, H. Zhang, T. Wang, K. Song, C. Zhu, J. Xu,Applied Energy2025,388, 125646, DOI 10. 1016/j.apenergy.2025.125646

    2025

  34. [42]

    Baumh¨ofer, M

    T. Baumh¨ofer, M. Br¨uhl, S. Rothgang, D. U. Sauer, Journal of Power Sources2014,247, 332–338, DOI 10.1016/j.jpowsour.2013.08.108

    2013 doi

  35. [43]

    S. J. Harris, D. J. Harris, C. Li,Journal of Power Sources2017,342, 589–597, DOI 10 . 1016 / j . jpowsour. 2016 . 12 . 083, http : / / dx . doi . org / 10 . 1016/j.jpowsour.2016.12.083

    2016

  36. [44]

    C. J. Ko, K. C. Chen,Applied Energy2024,357, 122488, DOI 10.1016/j.apenergy.2023.122488

    2023

  37. [45]

    A. A. Wang, S. Greenbank, G. Li, D. A. Howey, C. W. Monroe,Cell Reports Physical Science2022, 3, DOI 10.1016/j.xcrp.2022.101047. 20|1–22

    2022

  38. [46]

    Doyle, T

    M. Doyle, T. F. Fuller, J. Newman,Journal of The Electrochemical Society1993,140, 1526–1533, DOI 10.1149/1.2221597

    doi

  39. [47]

    Sulzer, S

    V. Sulzer, S. G. Marquis, R. Timms, M. Robinson, S. J. Chapman,Journal of Open Research Software 2021,9, 1–8, DOI 10.5334/JORS.309

    2021 doi

  40. [48]

    L. O. Valøen, J. N. Reimers,Journal of The Electro- chemical Society2005,152, A882, DOI 10.1149/1. 1872737

    doi

  41. [49]

    S. D. Talian, S. Brutti, M. A. Navarra, J. Moˇskon, M. Gaberscek,Energy Storage Materials2024,69, 103413, DOI 10.1016/j.ensm.2024.103413

    2024

  42. [50]

    Hallemans, N

    N. Hallemans, N. E. Courtier, C. P. Please, B. Planden, R. Dhoot, R. Timms, S. J. Chapman, D. Howey, S. R. Duncan,Journal of The Electrochem- ical Society2025, 1–20, DOI 10.1149/1945-7111/ add41b, http://arxiv.org/abs/2412.10896

    1945 arXiv

  43. [51]

    Single, B

    F. Single, B. Horstmann, A. Latz,Journal of Phys- ical Chemistry C2019,123, 27327–27343, DOI 10.1021/acs.jpcc.9b07389

    doi

  44. [52]

    G ¨unther, D

    J. G ¨unther, D. Wycisk, R. S. Dev, A. Fill, K. P. Birke, R. Moos, J. P. Schmidt, M. Oldenburger, Journal of The Electrochemical Society2025,172, 030525, DOI 10.1149/1945-7111/adbfc5, https: //iopscience.iop.org/article/10.1149/1945-7111/ adbfc5

    1945 doi

  45. [53]

    Y. Kuhn, B. Rana, M. Philipp, C. Schmitt, R. Sci- pioni, E. Flores, D. Kopljar, S. Clark, A. Latz, B. Horstmann,arXiv2025, http://arxiv.org/abs/ 2505.13566

    arXiv

  46. [54]

    Horstmann, F

    B. Horstmann, F. Single, A. Latz,Current Opinion in Electrochemistry2019,13, 61–69, DOI 10.1016/ j.coelec.2018.10.013, https://doi.org/10.1016/j. coelec.2018.10.013

    2018 doi

  47. [55]

    Ivers-Tiff´ee, A

    E. Ivers-Tiff´ee, A. Weber,Journal of the Ceramic Society of Japan2017,125, 193–201, DOI 10.2109/ jcersj2.16267

  48. [56]

    Y. Zhao, S. K¨ucher, A. Jossen,Electrochimica Acta 2022,432, 141174, DOI 10.1016/j.electacta.2022. 141174

    2022 doi

  49. [57]

    Barthelm´e, N

    S. Barthelm´e, N. Chopin,Journal of the American Statistical Association2014,109, 315–333, DOI 10.1080/01621459.2013.864178. 0 200 400 600 Elapsed time / h 0.2 0.4 0.6 0.8Voltage / V 0 20 40 60 80 Cycle number 0.04 0.02 0.00 0.02 0.04 Current / A Figure A-1: The GITT protocol ...

    2013

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.