Pith. sign in

REVIEW 4 major objections 7 minor 31 references

Combining CC and CV charging indicators beats either alone for battery health, and ordinary cross-validation overstates accuracy by about 119%.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-30 21:08 UTC pith:4VXGWNJD

load-bearing objection Solid NASA LOBO head-to-head of CC/CV indicators with a clear 5-fold vs LOBO warning; complementarity ranking is real but statistically thin on four folds. the 4 major comments →

arxiv 2607.23482 v1 pith:4VXGWNJD submitted 2026-07-26 cs.LG

Charging Phase Health Indicators for Battery State-of-Health Estimation: A Systematic Comparison of CC, CV, and Combined Approaches under Cross-Battery Validation

Huy Hoang Le , Kim-Anh Nguyen This is my paper
classification cs.LG
keywords battery state-of-healthcharging phase indicatorscross-battery validationlithium-ion batterymachine learningLeave-One-Battery-OutCC-CV chargingLightGBM
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper asks which simple features from a battery’s constant-current (CC) and constant-voltage (CV) charge phases actually predict State-of-Health when the model must work on a battery it has never seen. On the NASA aging cells, four CV-phase indicators plus CC duration are compared under Leave-One-Battery-Out validation. The combined CC+CV set wins (R² = 0.874), showing the two phases carry complementary aging signals—capacity fade in CC time, resistance and charge-acceptance changes in CV. The same models look far better under ordinary 5-fold cross-validation; LOBO RMSE is about 119% higher on average, so random splits overstate deployable accuracy. The authors turn that into practical selection rules: use both phases when full charge logs exist, fall back to CV-only when CC is incomplete or compute is tight, and treat ~4–5% RMSE as the realistic target for new cells.

Core claim

Under Leave-One-Battery-Out validation on four NASA LiCoO2 cells, the combined set of four CV-phase indicators plus CC phase duration achieves the best SOH estimates (R² = 0.874, RMSE 3.68%), beating CV-only (R² = 0.796) and CC-duration alone (R² = 0.845). That ranking shows CC and CV phases capture complementary degradation. Separately, LOBO RMSE averages about 119% higher than 5-fold cross-validation across models, so conventional splits substantially overestimate practical cross-battery accuracy.

What carries the argument

Leave-One-Battery-Out (LOBO) comparison of indicator sets: CV duration, CV-to-CC time ratio, current-decay time constant τ, CV charge throughput, and CC duration, scored with LightGBM (and checked against other gradient-boosting models) plus SHAP importance, which ranks the CV-to-CC ratio highest.

Load-bearing premise

That findings from four room-temperature NASA LiCoO2 18650 cells on one fixed CC–CV protocol are representative enough to guide indicator choice on other chemistries, temperatures, and BMS cutoff settings.

What would settle it

Repeat the same LOBO indicator-set comparison on another public aging set (different chemistry or CC–CV cutoffs); if combined CC+CV no longer beats CC-only and CV-only, or the CV-vs-LOBO gap collapses, the central ranking and the overestimation claim fail.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • When full CC–CV logs are available, prefer the five-indicator combined set for highest cross-battery SOH accuracy.
  • When only CV is logged or CC is adaptive/unstable, CV-only indicators remain usable without numerical differentiation.
  • Expect roughly 4–5% RMSE on unseen batteries of this type, not the ~2% suggested by random 5-fold CV.
  • SHAP ranking supports treating the dimensionless CV-to-CC time ratio as a primary, noise-robust health feature in BMS design.
  • Simple CV-duration thresholds can support field maintenance alarms without a full SOH model.

Where Pith is reading between the lines

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

  • If the 119% CV–LOBO gap is typical, many published sub-1% SOH numbers from mixed-cycle splits are not deployment-ready until re-checked with battery-held-out protocols.
  • The complementarity claim suggests multi-phase feature design may matter more than swapping among similar tree ensembles, which the paper’s model comparison already shows cluster tightly under LOBO.
  • A natural next test is whether the same CV-to-CC ratio stays top-ranked after temperature swings or after recalibrating the CV voltage threshold on CALCE/Oxford-style datasets.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The manuscript compares health indicators extracted from the constant-current (CC) and constant-voltage (CV) phases of CC–CV charging for battery state-of-health (SOH) estimation. Using four NASA LiCoO2 18650 cells (B0005/6/7/18) and LightGBM (plus XGBoost, CatBoost, Random Forest baselines), the authors evaluate four CV-phase indicators (CV duration, CV/CC time ratio, current-decay time constant, CV charge throughput) and CC duration, individually and combined, under Leave-One-Battery-Out (LOBO) validation. Headline results: the combined set achieves pooled R² = 0.874 (RMSE 3.68%) versus 0.845 for t_CC alone and 0.796 for CV-only (Table 7), supporting a complementarity claim; and LOBO RMSE averages ~119% higher than 5-fold CV (Table 9), quantifying how random splits overestimate deployable accuracy. SHAP analysis ranks the CV/CC time ratio as dominant, and practical indicator-selection guidelines (Table 11) are offered.

Significance. If the results hold, the paper makes a useful, practice-oriented contribution: it quantifies, on public and widely used NASA data, the gap between conventional random-split cross-validation and cross-battery evaluation — a message the SOH literature needs, since many published sub-1% RMSE claims rest on random splits. The work is careful in several respects that deserve explicit credit: a voltage-threshold sensitivity analysis (Appendix C), a nominal-capacity sensitivity check (§3.1), a hyperparameter robustness check (§3.4), per-battery error statistics (Appendix B), a model-parity analysis showing the bottleneck is generalization rather than model capacity (Table 6), and honest limitations including a negative chronological-split result (§5.5). The 98.9% CV-detection success rate and the simple, differentiation-free indicator extraction are genuine practical strengths. The complementarity claim itself, however, currently rests on pooled metrics over only four folds and needs stronger statistical support before the Table 11 guideline built on it can be considered established.

major comments (4)
  1. [§4.4, Table 7] The central complementarity claim — Combined R²=0.874 > t_CC-only 0.845 > CV-only 0.796 — rests on pooled LOBO metrics over only four folds, with no per-fold indicator-set breakdown. The paper's own Table 5 shows large per-fold heterogeneity (per-battery R² from 0.609 to 0.896). With N=4 folds, a ΔR² of 0.029 between Combined and t_CC-only could be carried by a single fold (B0006 or B0018). The authors should report per-fold metrics for each indicator set in Table 7, plus a paired comparison across folds (e.g., per-fold ΔRMSE with sign consistency), or temper the complementarity conclusion and the 'Complete CC-CV data available → Combined' recommendation in Table 11 accordingly.
  2. [§4.3.2/§4.4, Tables 5 and 7] The 'Overall R²' values appear to be computed by pooling predictions across folds, where each fold's predictions come from a differently-trained model. Pooled R² credits the model with between-battery SOH variance, which is an easier task than within-battery degradation tracking — the decision-relevant quantity for the maintenance use case the paper targets. The manuscript's own numbers reveal the discrepancy: Table 5 reports per-fold R² = 0.769 ± 0.110 while the text and Table 7 quote 0.796 for the same CV-only configuration. The authors should (i) state explicitly how 'Overall R²' is computed, (ii) report both pooled and mean per-fold (within-battery) R² for all indicator sets in Table 7, and (iii) verify that the Combined > t_CC-only ordering survives under the within-battery metric.
  3. [§3.2 and Appendix C vs. §4.3.2, Table 5] Numerical inconsistency: the threshold sensitivity analysis (§3.2, Table C1) reports the baseline 4.17 V configuration at LOBO RMSE = 4.235% and R² = 0.808, but Table 5 reports the same CV-only LOBO configuration at RMSE = 4.69% and R² = 0.796. These should be identical experiments. Please explain the discrepancy (different model configuration? different random seed? pooled vs. mean-per-fold computation?) and reconcile the two tables, since Appendix C is the basis for the robustness claim about threshold choice.
  4. [§3.4] Feature standardization is described as 'z-score normalization before model training' without stating whether the scaler is fit on training folds only. If statistics are computed over all 623 samples before the LOBO split, test-battery information leaks into every fold. The effect is likely small for standardization, but given the paper's central message is evaluation rigor, the pipeline (fit scaler within each training fold, apply to test fold) should be stated explicitly and, if necessary, corrected.
minor comments (7)
  1. [§3.1 vs. Table 1] Direct contradiction on SOH>100% cycles: the text states these are 'concentrated in B0006 (18 cycles, max 104.6%) and B0007 (20 cycles, max 101.7%)', but Table 1 lists B0006's SOH range as 69.8–99.6 and assigns the 104.6% maximum to B0018. Please correct whichever is wrong.
  2. [§3.3, Indicator 3] Cycle-count inconsistency: the text states exponential fitting converged for '636 of 637 CV cycles', but §4.1 reports 623 valid cycles out of 630 charge cycles. The 637/630 discrepancy (and per-battery counts summing to 637) should be reconciled.
  3. [Table 2] t_CV/t_CC and τ are both reported with r = -0.719 to three decimals. If this is a coincidence of rounding, fine, but given that τ is derived from the same current-decay profile, a note on the near-collinearity of the CV indicators (and its implications for the CV-only set) would strengthen the analysis.
  4. [§4.5, Table 8] Two different SHAP summaries are given (mean |SHAP| = 7.839 for the ratio in Table 8; 6.77 ± 1.56 'under LOBO validation' in the text) without stating which model/validation produces the Table 8 values. Please clarify; presumably Table 8 is from a model trained on all data or under 5-fold CV.
  5. [§4.3.1/§3.4] The hyperparameter grid search is evaluated with 5-fold CV and the default configuration retained; this is reasonable, but note that selecting hyperparameters on 5-fold CV while reporting LOBO as primary is methodologically slightly mismatched. A sentence acknowledging this would suffice.
  6. [§5.5] The limitations section is commendably honest. Consider adding that with only four LOBO folds, the ±1.26% RMSE spread and the 119% CV-to-LOBO gap are themselves noisy estimates; cross-dataset replication (already listed as future work) is the natural remedy.
  7. [General] No code or extracted-indicator data availability statement is given. Given the public NASA dataset and the reproducibility-oriented framing, releasing the extraction and evaluation code would substantially increase the paper's usefulness.

Circularity Check

0 steps flagged

No circularity: SOH labels are independent discharge measurements; CC/CV indicators are inputs to a supervised fit, not definitional of the target.

full rationale

The paper is a standard empirical comparison of hand-crafted charging-phase features under LOBO validation. SOH is defined from discharge coulomb counting relative to a fixed nominal capacity (Section 3.1), while the predictors (t_CV, t_CV/t_CC, τ, Q_CV, t_CC) are extracted solely from the charge curve (Sections 3.2–3.3). Nothing in the feature definitions algebraically forces the LightGBM (or RF/XGBoost/CatBoost) outputs to match SOH; the reported R²/RMSE values are ordinary out-of-sample supervised metrics. SHAP rankings and Pearson correlations are post-hoc descriptions of the fitted model, not self-defining identities. The 119% CV-vs-LOBO gap is arithmetic from Table 9, not a derived law. There is no load-bearing self-citation, uniqueness theorem, or ansatz smuggled in via prior author work. Methodological choices (4.17 V threshold, 20 mA cutoff, default LightGBM hyperparameters) affect feature extraction but do not make predictions true by construction. The derivation chain is therefore self-contained against external benchmarks; circularity score is 0.

Axiom & Free-Parameter Ledger

5 free parameters · 5 axioms · 0 invented entities

Load-bearing content is empirical ML on a public aging set. The claim rests on standard SOH labeling, a fixed CC–CV protocol, hand-chosen detection thresholds, a small set of engineered indicators, and a default gradient-boosting configuration—not on new physical entities. Generalization from four lab cells is the main unproven bridge to the deployment guidelines.

free parameters (5)
  • CV detection voltage threshold = 4.17 V
    Transition marked when V first exceeds 4.17 V (not the 4.2 V setpoint); sensitivity table shows small metric movement but the operating point is chosen by authors.
  • Nominal capacity for SOH = 1.86 Ah
    SOH = measured discharge capacity / nominal; authors set nominal to observed maximum 1.86 Ah (vs manufacturer 2.0 Ah), which rescales RMSE though they report identical R².
  • LightGBM hyperparameters = 100 / depth 6 / lr 0.1
    Default-ish config (n_estimators=100, max_depth=6, lr=0.1) retained after a 27-point grid found a slightly better CV setting; model capacity is a free modeling choice.
  • Exponential τ model and fallback = NLLS + 36.8% fallback
    I(t)=I0 exp(−t/τ)+I_offset via NLS; 0.2% of cycles use 36.8%-of-initial-current time when fit fails—definitional choice affecting one indicator.
  • CV current cutoff and min points = 20 mA, ≥20 points
    CV end at 20 mA and ≥20 samples with decreasing current; defines which cycles and durations enter the study (623/630 kept).
axioms (5)
  • domain assumption Discharge coulomb-counted capacity over a fixed nominal is the ground-truth SOH label for supervised learning.
    Section 3.1 defines SOH from NASA discharge capacity; all R²/RMSE claims are relative to this label.
  • domain assumption CC–CV charging at the NASA protocol (1.5 A to 4.2 V, then CV to 20 mA) yields health-informative phase timings and current decay.
    Entire indicator pipeline assumes this protocol structure (§3.1–3.3); authors note other strategies may differ (§5.5).
  • ad hoc to paper Leave-One-Battery-Out on four cells is a sufficient proxy for realistic cross-battery deployment error.
    Primary validation design (§3.4, §4.3); with N=4 folds the “realistic expectations” language and guidelines lean on this proxy.
  • domain assumption Tree ensemble regression on z-scored tabular indicators is an adequate estimator for comparing indicator sets (model class not the bottleneck).
    Supported internally by similar LOBO scores across LightGBM/XGBoost/CatBoost/RF (Table 6), but still an untested assumption vs physics models or other feature types.
  • standard math Standard statistical learning and Pearson/SHAP interpretation tools apply to these cycle-level samples.
    Ordinary ML evaluation machinery; no exotic proof obligations.

pith-pipeline@v1.2.0-grok45-kimik3 · 19229 in / 3763 out tokens · 76805 ms · 2026-07-30T21:08:23.454226+00:00 · methodology

0 comments
read the original abstract

Accurate State-of-Health estimation is essential for safe battery operation and cost-effective maintenance. Although numerous health indicators have been derived from constant-current (CC) and constant-voltage (CV) charging phases, their effectiveness under realistic cross-battery validation remains insufficiently studied. This work addresses this gap through a systematic comparison of CC-only, CV-only, and combined indicator sets using rigorous Leave-One-Battery-Out (LOBO) validation on the NASA battery aging dataset. Four CV-phase indicators and CC phase duration are evaluated individually and in combination. Results show that the combined CC+CV approach achieves the best performance (R2 = 0.874), confirming that CC and CV phases capture complementary degradation information. Moreover, a 119% performance gap is observed between standard 5-fold cross-validation and LOBO validation, indicating that conventional evaluation overestimates practical accuracy. Based on these findings, practical guidelines are provided for indicator selection under data and computational constraints.

Figures

Figures reproduced from arXiv: 2607.23482 by Huy Hoang Le, Kim-Anh Nguyen.

Figure 1
Figure 1. Figure 1: Capacity degradation curves for the four batteries (B0005, B0006, B0007, and B0018). The charging protocol consists of CC charging at 1.5 A until the voltage reaches 4.2 V, followed by CV charging at 4.2 V until the current drops to 20 mA. Discharge is performed at 2 A constant current until the voltage reaches 2.7 V, with capacity [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Example of CV phase detection in a charging cycle. 3.3. Health indicator extraction Four health indicators are extracted from the CV charging phase and defined as follows: Indicator 1: CV phase duration, denoted 𝒕CV. The time duration of the CV charging phase in seconds, [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Correlation analysis of CV phase indicators: (a) Correlation heatmap showing relationships between all indicators and SOH; (b–e) Scatter plots of key indicators versus SOH, colored by battery ID, demonstrating consistent degradation trends across different batteries [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Evolution of CV phase duration over battery cycling The evolution of CV phase duration over battery cycling is analyzed to assess monotonicity [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: SOH prediction results under LOBO validation for all four batteries [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: SHAP indicator importance analysis showing mean absolute SHAP value [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Actual versus predicted SOH for all batteries under LOBO validation. [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗

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

31 extracted references · 16 canonical work pages

  1. [1]

    The development and future of lithium ion batteries

    Blomgren G E. The development and future of lithium ion batteries. Journal of The Electrochemical Society 2017; 164(1): A5019–A5025. https://doi.org/10.1149/2.0251701jes

    work page doi:10.1149/2.0251701jes 2017

  2. [2]

    Critical review of state of health estimation methods of Li-ion batteries for real applications

    Berecibar M, Gandiaga I, Villarreal I, Omar N, Van Mierlo J, Van den Bossche P. Critical review of state of health estimation methods of Li-ion batteries for real applications. Renewable and Sustainable E nergy Reviews 2016; 56: 572–587. https://doi.org/10.1016/j.rser.2015.11.042

    doi:10.1016/j.rser.2015.11.042 2016

  3. [3]

    A review of state of health and remaining useful life estimation methods for lithium-ion battery in electric vehicles: challenges and recommendations

    Lipu M S H, Hannan M A, Hussain A, Hoque M M, Ker P J, Saad M H M, Ayob A. A review of state of health and remaining useful life estimation methods for lithium-ion battery in electric vehicles: challenges and recommendations. Journal of Cleaner Production 2018; 205: 115–133. https://doi.org/10.1016/j.jclepro.2018.09.065

    work page doi:10.1016/j.jclepro.2018.09.065 2018

  4. [4]

    Battery lifetime prognostics

    Hu X, Xu L, Lin X, Pecht M. Battery lifetime prognostics. Joule 2020; 4(2): 310–346. https://doi.org/10.1016/j.joule.2019.11.018

    doi:10.1016/j.joule.2019.11.018 2020

  5. [5]

    A review on machinery diagnostics and prognostics implementing condition -based maintenance

    Jardine AKS, Lin D, Banjevic D. A review on machinery diagnostics and prognostics implementing condition -based maintenance. Mechanical Systems and Signal Processing 2006; 20(7): 1483–1510. https://doi.org/10.1016/j.ymssp.2005.09.012

    doi:10.1016/j.ymssp.2005.09.012 2006

  6. [6]

    Charging protocols for lithium-ion batteries and their impact on cycle life —an experimental study with different 18650 high-power cells

    Keil P, Jossen A. Charging protocols for lithium-ion batteries and their impact on cycle life —an experimental study with different 18650 high-power cells. Journal of Energy Storage 2016; 6: 125–141. https://doi.org/10.1016/j.est.2016.02.005

    work page doi:10.1016/j.est.2016.02.005 2016

  7. [7]

    Battery Management Systems

    Plett G L. Battery Management Systems. V olume I: Battery Modeling. Artech House; 2015

    2015

  8. [8]

    Identify capacity fading mechanism in a commercial LiFePO4 cell

    Dubarry M, Liaw B Y . Identify capacity fading mechanism in a commercial LiFePO4 cell. Journal of Power Sources 2009; 194(1): 541–

    2009

  9. [9]

    On-board state of health monitoring of lithium -ion batteries using incremental capacity analysis with support vector regression

    Weng C, Cui Y , Sun J, Peng H. On-board state of health monitoring of lithium -ion batteries using incremental capacity analysis with support vector regression. Journal of Power Sources 2013; 235: 36–44. https://doi.org/10.1016/j.jpowsour.2013.02.012

    work page doi:10.1016/j.jpowsour.2013.02.012 2013

  10. [10]

    Gaussian process regression for forecasting battery state of h ealth

    Richardson R R, Osborne M A, Howey D A. Gaussian process regression for forecasting battery state of h ealth. Journal of Power Sources 2017; 357: 209–219. https://doi.org/10.1016/j.jpowsour.2017.05.004

    work page doi:10.1016/j.jpowsour.2017.05.004 2017

  11. [11]

    Data-driven prediction of battery cycle life before capa city degradation

    Severson K A, Attia P M, Jin N, Perkins N, Jiang B, Yang Z, Chen M H, Aykol M, Herring P K, Fraggedakis D, Bazant M Z, Harris S J, Chueh W C, Braatz R D . Data-driven prediction of battery cycle life before capa city degradation. Nature Energy 2019; 4(5): 383–391. https://doi.org/10.1038/s41560-019-0356-8

    doi:10.1038/s41560-019-0356-8 2019

  12. [12]

    Long short-term memory recurrent neural network for remaining useful life prediction of lithium - ion batteries

    Zhang Y , Xiong R, He H, Pecht M G. Long short-term memory recurrent neural network for remaining useful life prediction of lithium - ion batteries. IEEE Transactions on Vehicular Technology 2018; 67(7): 5695–5705. https://doi.org/10.1109/TVT.2018.2805189

    arXiv 2018

  13. [13]

    A deep learning method for online capacity estimation of lithium -ion batteries

    Shen S, Sadoughi M, Chen X, Hong M, Hu C. A deep learning method for online capacity estimation of lithium -ion batteries. Journal of Energy Storage 2019; 25: 100817. https://doi.org/10.1016/j.est.2019.100817

    arXiv 2019

  14. [14]

    Convolutional Neural Network - Gated Recurrent Unit combined with Error Correction for Lithium Battery State of Health Estimation

    Luo X, Bu W, Liang H, Zheng M. Convolutional Neural Network - Gated Recurrent Unit combined with Error Correction for Lithium Battery State of Health Estimation. Eksploatacja i Niezawodnos c – Maintenance and Reliability 2025; 27(4): 202184. https://doi.org/10.17531/ein/202184

    work page doi:10.17531/ein/202184 2025

  15. [15]

    Application of DBN -based KRL S method for RUL prediction of lithium -ion batteries

    Li J, Ding P. Application of DBN -based KRL S method for RUL prediction of lithium -ion batteries. Eksploatacja i Niezawodnos c – Maintenance and Reliability 2025; 27(2): 194174. https://doi.org/10.17531/ein/194174

    work page doi:10.17531/ein/194174 2025

  16. [16]

    A modified TimeGAN-based data augmentation ap proach for the state of health prediction of Lithium-Ion Batteries

    Echabarri S, Do P, Vu H C, Liegeois P Y . A modified TimeGAN-based data augmentation ap proach for the state of health prediction of Lithium-Ion Batteries. Reliability Engineering and System Safety 2025; 264(Part A): 111297. https://doi.org/10.1016/j.ress.2025.111297

    arXiv 2025

  17. [17]

    Machine learning pipeline for battery state -of-health estimati on

    Roman D, Saxena S, Robu V , Pecht M, Flynn D. Machine learning pipeline for battery state -of-health estimati on. Nature Machine Intelligence 2021; 3: 447–456. https://doi.org/10.1038/s42256-021-00312-3. Eksploatacja i Niezawodność – Maintenance and Reliability V ol. 28, No. 4, 2026

    doi:10.1038/s42256-021-00312-3 2021

  18. [18]

    Predicting the state of charge and health of batter ies using data-driven machine learni ng

    Ng M F, Zhao J, Yan Q, Conduit G J, Seh Z W. Predicting the state of charge and health of batter ies using data-driven machine learni ng. Nature Machine Intelligence 2020; 2: 161–170. https://doi.org/10.1038/s42256-020-0156-7

    work page doi:10.1038/s42256-020-0156-7 2020

  19. [19]

    A quick on-line state of health estimation method for Li -ion battery with incremental capacity curves processed by Gaussian f ilter

    Li Y , Abdel-Monem M, Gopalakrishnan R, Berecibar M, Nanini-Maury E, Omar N, Van den Bossche P, Van Mierlo J. A quick on-line state of health estimation method for Li -ion battery with incremental capacity curves processed by Gaussian f ilter. Journal of Power Sources 2018; 373: 40–53. https://doi.org/10.1016/j.jpowsour.2017.10.092

    work page doi:10.1016/j.jpowsour.2017.10.092 2018

  20. [20]

    Multi -kernel relevance vector machine with parameter optimization for cycling aging prediction of lithium-ion batteries

    Jiang B, Dai H, Wei X, Xu L, Xu Z. Multi -kernel relevance vector machine with parameter optimization for cycling aging prediction of lithium-ion batteries. IEEE Journal of Emerging and Selec ted Topics in Power Electronics 2023; 11(1): 175–186. https://doi.org/10.1109/JESTPE.2021.3133697

    arXiv 2023

  21. [21]

    Prediction of remaining useful life for lithium -ion battery with multiple health indicators

    Su C, Chen H, Wen Z. Prediction of remaining useful life for lithium -ion battery with multiple health indicators. Eksploatacja i Niezawodnosc – Maintenance and Reliability 2021; 23(1): 176–183. https://doi.org/10.17531/ein.2021.1.18

    work page doi:10.17531/ein.2021.1.18 2021

  22. [22]

    Useful energy prediction mo del of a Lithium -ion cell operating on various duty cycles

    Burzynski D. Useful energy prediction mo del of a Lithium -ion cell operating on various duty cycles. Eksploatacja i Niezawodnos c – Maintenance and Reliability 2022; 24(2): 317–329. https://doi.org/10.17531/ein.2022.2.13

    work page doi:10.17531/ein.2022.2.13 2022

  23. [23]

    A study of the relationship betwe en coulombic efficiency and capacity degradation of commercial lithium-ion batteries

    Yang F, Wang D, Zhao Y , Tsui K L, Bae S J. A study of the relationship betwe en coulombic efficiency and capacity degradation of commercial lithium-ion batteries. Energy 2018; 145: 486–495. https://doi.org/10.1016/j.energy.2017.12.144

    work page doi:10.1016/j.energy.2017.12.144 2018

  24. [24]

    Online state of health estimation for lithium-ion batteries based on support vector machine

    Chen Z, Sun M, Shu X, Xiao R, Shen J. Online state of health estimation for lithium-ion batteries based on support vector machine. Applied Sciences 2018; 8(6): 925. https://doi.org/10.3390/app8060925

    work page doi:10.3390/app8060925 2018

  25. [25]

    Differential current in constant -voltage charging mode: a novel tool for state -of-health and state -of-charge estimation of lithium-ion batteries

    Ko C J, Chen K C, Su T W. Differential current in constant -voltage charging mode: a novel tool for state -of-health and state -of-charge estimation of lithium-ion batteries. Energy 2024; 288: 129826. https://doi.org/10.1016/j.energy.2023.129826

    arXiv 2024

  26. [26]

    Battery data set, NASA Ames Prognostics Data Repository

    Saha B, Goebel K. Battery data set, NASA Ames Prognostics Data Repository. NASA Ames Research Center; 2007. Available from: https://www.nasa.gov/content/prognostics-center-of-excellence-data-set-repository

    2007

  27. [27]

    LightGBM: a highly efficient gradient boosting decision tree

    Ke G, Meng Q, Finley T, Wang T, Chen W, Ma W, Ye Q, Liu T-Y . LightGBM: a highly efficient gradient boosting decision tree. Advances in Neural Information Processing Systems 2017; 30: 3146–3154. https://proceedings.neurips.cc/paper/2017/file/6449f44a102fde848669bdd9eb6b76fa -Paper.pdf

    2017

  28. [28]

    A unified approach to interpreting model predictions

    Lundberg S M, Lee S I. A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems 2017; 30: 4765–4774. https://proceedings.neurips.cc/paper/2017/file/8a20a8621978632d76c43dfd28b67767-Paper.pdf

    2017

  29. [29]

    A data-driven predictive maintenance strategy based on accurate failure prognostics

    Chen C, Wang C, Lu N, Jiang B, Xing Y . A data-driven predictive maintenance strategy based on accurate failure prognostics. Eksploatacja i Niezawodnosc – Maintenance and Reliability 2021; 23(2): 387–394. https://doi.org/10.17531/ein.2021.2.19

    work page doi:10.17531/ein.2021.2.19 2021

  30. [30]

    Application of machine learning and rough set theory in lean maintenance decision support system development

    Antosz K, Jasiulewicz -Kaczmarek M, Pasko L, Zhang C, Wang S. Application of machine learning and rough set theory in lean maintenance decision support system development. Eksploatacja i Niezawodnos c – Maintenance and Reliability 2021; 23(4): 695–708. https://doi.org/10.17531/ein.2021.4.15. Appendix A: Detailed indicator statistics Table A1 presents the ...

    work page doi:10.17531/ein.2021.4.15 2021

  31. [549]

    https://doi.org/10.1016/j.jpowsour.2009.05.036

    work page doi:10.1016/j.jpowsour.2009.05.036 2009