Long-term degradation of vanadium redox flow batteries (VRFBs) is strongly coupled with the evolution of discharge direct current internal resistance (DCIR). This work employs 213 constant-current cycles on a 4 cm2 single cell to examine the associations between discharge DCIR and capacity, discharge voltage and efficiencies. The average values of the initial 10 cycles were used as the baseline. Multiple statistical approaches, including detrending, differencing, moving-block bootstrap, and internal chronological holdout evaluation were used to assess the influence of shared temporal trends and serial dependence. Self-calculated DCIR matches test records closely with merely 1.03% average relative error. After trend correction, normalized DCIR maintains a strong negative correlation with normalized capacity. The established free-intercept quadratic model achieved the best full-data fitting performance, with an R2 of 0.99593 and a root mean square error (RMSE) of 0.00524. The fitted empirical relationship indicates that equal increments in normalized DCIR are associated with larger concurrent capacity-state reductions in the higher-resistance region. During the internal chronological holdout evaluation, the DCIR-based model yielded an RMSE of 0.00854. These results establish a statistically robust DCIR–capacity relationship over long-term cycling and demonstrate the potential of routinely recorded discharge DCIR as a low-cost concurrent capacity-state indicator, providing a quantitative foundation for its future extension to broader VRFB operating scenarios.
Accurate and robust state-of-health (SOH) estimation is essential for efficient battery management systems (BMS), especially in practical scenarios suffering from severe cell-to-cell inconsistencies and data distribution shifts. Although prevailing data-driven estimation methods can achieve satisfactory accuracy within specific domains, their generalization capability degrades drastically when applied to unseen battery cells. To fill this research gap, this paper develops an integrated physics-guided and domain-aware SOH estimation framework. The proposed framework combines redundancy-aware feature screening, cycle-aware soft covariance alignment specifically designed for physics-based health indicators, as well as monotonicity-constrained LightGBM regression embedded with split conformal uncertainty quantification. Experimental validations are conducted on the NASA B0005, B0006 and B0007 battery datasets under the leave-one-cell-out (LOCO) cross-cell evaluation strategy. Comparative results reveal that the presented method achieves prominent performance improvement on the most difficult B6 domain, where the root mean square error (RMSE) is reduced from 0.1068 to 0.0845 with a decline rate of 20.9%, and the coefficient of determination (R2) rises from 0.2464 to 0.5286, while maintaining stable estimation accuracy on less challenging target domains. In terms of overall performance, the average RMSE across all tested cells drops from 0.0499 to 0.0425, corresponding to a 14.8% reduction, indicating improved cross-cell performance across the investigated cells. Furthermore, the adaptive alignment mechanism can be dynamically activated only when necessary, effectively avoiding redundant distortion of the original feature distribution. The research findings indicate that the integrated framework can alleviate cross-cell battery degradation discrepancies under the investigated conditions. The proposed strategy demonstrates promising potential for improving cross-cell SOH estimation under the investigated laboratory conditions; however, the uncertainty intervals are not fully calibrated under severe domain shift, and broader validation on larger and more heterogeneous battery datasets is required before practical deployment claims.
Battery-management systems observe current, terminal voltage, limited temperature measurements, and operating setpoints, but not the internal variables of electrochemical theory, so routine signals generally cannot identify a unique mechanism. We present a human-guided, generative-AI-assisted methodology that translates physics-based expectations into deterministic, auditable screening rules: human scientific authority fixes the physical assumptions, evidence requirements, and permissible claims, artificial intelligence supports development, and runtime evaluation is non-generative. LiFePO4 is the test case. A Zeng–Bazant current-dependent plateau approximation supplies a physics-based reference, and a bivariate representational precedent motivates a composite, reference-dependent voltage residual that is not identified as thermodynamic work. Eleven observable screens return present, absent within resolution, indeterminate, or unavailable. Four evidence forms are separated. Digitized model curves show the reduced plateau relation tracks its parent phase-field simulation through moderate rates, with a high-rate limitation. Published temperature-conditioned discharge profiles show stable plateau elevation and flattening from 268 to 298 K across 0.5 C–2 C, with mixed 2 C curvature. Measured replicates of a commercial cylindrical cell at two ambient setpoints resolve a within-run surface-temperature depression whose integrated first-law balance is heat-rejection-dominant and compatible with, but not uniquely attributed to, a literature-bounded reversible contribution. Controlled synthetic cases verify deterministic feature recovery and abstention without establishing a mechanism.
Solid-state sodium metal batteries are promising for large-scale energy storage owing to their high safety, abundant sodium resources, and low material cost. However, NASICON-type solid-state electrolytes, such as Na3Zr2Si2PO12, still suffer from limited room-temperature ionic conductivity and poor ceramic densification. Herein, a multivalent co-doping strategy using Zn2+, Sc3+, Hf4+, and Nb5+ was employed to regulate the crystal structure, sintering behavior, and Na+ transport properties of NASICON electrolytes. The effects of isovalent and aliovalent dopants were systematically investigated by XRD, Rietveld refinement, SEM-EDS, bond-valence-site calculations, and electrochemical impedance spectroscopy. The optimized monoclinic Na3.167Zn0.167Hf0.167Zr1.5Nb0.167Si2PO12 electrolyte delivered a room-temperature total ionic conductivity of 1.16 mS cm−1 and an activation energy of 0.35 eV, mainly due to optimized Na+ migration-channel geometry and enhanced ceramic densification. Na||Na symmetric cells exhibited stable cycling for 500 h with a maximum polarization voltage of 20 mV. Furthermore, solid-state Na||Na3V2(PO4)3 cells retained 95% capacity after 624 cycles at 1 C. This work provides a feasible doping strategy for advanced NASICON electrolytes and solid-state sodium metal batteries.
This study involved connecting four 18650 cylindrical lithium-ion batteries in series. An ABS structure produced using a 3D printer was then filled with salt hydrate phase change material (PCM). The batteries were then tested in two separate systems: one containing PCM and one without. In the PCM-enhanced system, the maximum surface temperature did not exceed 46 °C during discharge at a constant current of 8 A, corresponding to a discharge rate of 2.85 C. In the PCM-free system, the maximum surface temperature reached 66.3 °C under the same discharge condition. This resulted in improved battery performance by reducing temperature rise and enhancing discharge behavior under high-rate operating conditions. Furthermore, using PCM increased the batteries’ energy output by 20% during discharge at a constant 2.85 C rate. Finally, a techno-economic and environmental analysis was conducted to assess the system’s suitability for micro-mobility applications with a 1 kWh capacity. The system reached the break-even point in 4 years and 2 months, indicating that it is economical and feasible.
This work develops a hybrid kernel sparse Gaussian process regression integrated with kernel density estimation (HCSGPR-UQ) to resolve three critical drawbacks of conventional lithium-ion battery State of Energy (SOE) estimators: degraded accuracy under dynamic loads, high computational overhead, and inadequate uncertainty quantification. A composite covariance kernel is built by weighting the radial basis function (RBF) and Matérn 5/2 kernels to simultaneously model global smooth SOE decay trends and local nonlinear fluctuations induced by abrupt current/temperature variations. Inducing-point sparse approximation is adopted to accelerate model inference, while kernel density estimation (KDE) generates nonparametric prediction bounds for quantitative uncertainty evaluation. Validations are carried out on a 75 Ah traction lithium-ion cell across −5 °C to 35 °C under Dynamic Stress Test (DST) and Beijing Bus Dynamic Stress Test (BBDST) cycles. Experimental results reveal that the proposed method yields mean absolute errors (MAEs) of only 0.32% (DST) and 0.38% (BBDST), runs roughly 15× faster than full Gaussian process regression (GPR), and attains a 94.7% coverage probability for nominal 95% prediction intervals. Balancing estimation precision, real-time inference speed and statistical reliability, the proposed framework delivers a viable online SOE estimation solution for vehicle battery management systems (BMSs).
With the widespread application of lithium-ion batteries in electric vehicles, degradation diagnosis has attracted increasing attention. In practical operating scenarios, however, path-dependent degradation induced by the alternating effects of calendar aging and cycling aging can significantly influence the diagnosis of battery degradation. Existing methods often identify degradation under these two aging conditions in isolation, making it difficult to quantify their coupled impact. To address this issue, this study applies a physically constrained half-cell OCP reconstruction framework to quantify electrode-level degradation under coupled aging conditions. Specifically, degradation parameters are identified by fitting full-cell pseudo-open-circuit voltage (pOCV) curves with half-cell open-circuit potential (OCP) profiles, and the corresponding degradation modes are further quantified. The results show that the proposed model can reconstruct full-cell pOCV under different aging conditions with an RMSE maintained below 10.5 mV. The loss of active material in the anode is highly sensitive to continuous cycling, whereas the divergence of loss of lithium inventory under alternating aging conditions is relatively weak but shows pronounced differences under distinct single-aging conditions. This method enables electrode-level diagnosis of path-dependent degradation under alternating aging conditions and provides a foundation for reliable degradation diagnosis under complex operating scenarios.
The electrification of commercial vehicles demands precise battery thermal management, but direct measurement of the cell core temperature is challenging. This paper presents an electrochemical impedance spectroscopy (EIS)-based approach for rapid indirect estimation of the mean internal temperature in 2170 NMC lithium-ion cells. Three measurement approaches and various fitting methods, including Steinhart–Hart, least-squares polynomials, and nonlinear Arrhenius-based fits, are compared using experimental data. The results indicate that estimation accuracy is more strongly influenced by the selection of measurement frequency than by the choice of fitting approach. The optimal method combines a single optimized frequency with a nonlinear polynomial incorporating an Arrhenius term, achieving a maximum deviation of 0.23K and a mean deviation of 0.14K. This framework enables indirect EIS-based estimation of the cell core temperature and can be further refined through measurements on multiple cells or by combining different fitting methods. Future work should extend the proposed approach to dynamic EIS measurements, battery ageing, and integration with thermal models for early overheating warning and identification of the first thermal runaway warning stage in high-power commercial vehicles.
Fault detection from electric vehicle charging segments is challenging because real-world records are noisy, verified fault labels are scarce, and weak signatures may evolve gradually across time and unevenly across a vehicle’s charging history. This study proposes an unsupervised dynamic multiscale state-space support vector data description framework for vehicle-level battery fault detection. A gated diagonal state-space encoder preserves long-range charging patterns while retaining a direct input path for transient changes. A progressive cross-scale fusion module then combines short-term fluctuations with accumulated deviations in the learned hidden representation. Finally, a two-stage hypersphere optimisation strategy first estimates the normal centre and then refines the boundary around that fixed centre. This coordinated design avoids sequence reconstruction and directly scores charging segments by their distance from normal behaviour before robust vehicle-level aggregation. On the two EVBattery subsets that permit statistically reliable evaluation, the proposed framework achieved vehicle-level areas under the receiver operating characteristic curves of 0.8849 and 0.8438. These results exceed those of the best-performing baseline on the corresponding subsets by 0.0889 and 0.0417, respectively. The results show that coordinating long-range encoding, cross-scale fusion, and staged boundary learning improves threshold-independent vehicle-level fault ranking in real charging data.
Solid electrolyte interphase (SEI) characterization is needed to interpret the performance, degradation, and lifetime of graphite and Si-containing carbon-based negative electrodes in lithium-ion batteries. However, SEI claims are often difficult to compare because measured signals, inferred assignments, sample history, and electrode architecture are not always clearly separated. This review presents a claim-bounded framework for SEI characterization that distinguishes direct observables from inferred chemical, molecular, structural, morphological, and functional information. Photoelectron spectroscopy methods provide chemical-state and relative-depth-sensitivity constraints; secondary-ion mass spectrometry methods provide fragment and isotope distributions; vibrational spectroscopies support functional-group and local vibrational evidence; nuclear magnetic resonance and molecular mass spectrometry provide molecular or product-level constraints; and microscopy, tomography, and atomic force microscopy provide morphology, architecture, local thickness, topography, and mechanical response. Across these methods, rinsing, drying, sputtering, beam exposure, extraction, and limited sampling can alter the observable and therefore the defensible claim. The review emphasizes the distinction between native electrode-associated SEI features and extracted, soluble, or electrolyte-phase products, and between morphology-only evidence and chemically assigned morphology. It concludes by proposing claim-driven correlative workflows and reporting guidance for reproducible interpretation on graphite, Si/graphite, Si/C, and carbon-coated Si architectures where directly studied or present.
Lithium-ion battery fires are governed by continuous heat release during thermal runaway, flammable gas venting, and thermal coupling between adjacent cells. Even after visible flames are extinguished, post-extinguishment temperature rise, thermal runaway propagation, and reignition may still occur. This review establishes a full-process control framework linking flame suppression, sustained cooling, thermal runaway propagation mitigation, and reignition control. Within this framework, water-based agents, clean gaseous agents, dry powder agents, foams, cryogenic media, and hybrid suppression methods are not only compared by their flame extinguishment performance but also by their cooling capability, thermal runaway propagation/reignition control, scenario applicability, and environmental impacts. Evaluation metrics and standardization requirements are further integrated to support cross-study comparison and practical suppressant selection. Existing studies indicate that a single suppressant is generally unable to achieve both rapid flame extinguishment and post-extinguishment thermal stability. Multi-mechanism synergy, realistic scenario validation, and standardized evaluation protocols are therefore essential for improving the full-process control of lithium-ion battery fires.
Lithium metal batteries are widely regarded as one of the most promising candidates for achieving energy densities beyond 500 Wh kg−1. However, their practical commercialization is severely hindered by the high reactivity of lithium metal, which leads to pronounced interfacial instability. Constructing an artificial solid electrolyte interphase (SEI) via surface pretreatment has been demonstrated to be an effective strategy for suppressing dendrite growth and mitigating parasitic side reactions. Herein, we take advantage of the rapid reaction between lipoic acid (LA) and lithium metal to pre-form a uniform artificial SEI on the anode surface. This interphase is composed of organic COO-Li species and sulfur-containing compounds. Electrochemically, the LA-modified lithium anode exhibits remarkable stability, sustaining more than 1500 h of cycling in symmetric cells at 5 mA cm−2 and 5 mAh cm−2. Furthermore, full-cell configurations, including lithium-sulfur and lithium-LiFePO4 pouch cells, deliver significantly improved cycling performance compared with those employing bare lithium anodes. These results establish a practical and scalable route for fabricating stable artificial SEI layers on lithium metal, thereby providing a feasible pathway toward the realization of high-energy-density lithium metal batteries.
Sodium-ion batteries are emerging as credible complements to lithium-ion technology for sustainable, safe, and cost-effective energy storage. This critical review links molecular-scale electrolyte solvation and interphase chemistry to electrode compatibility, full-cell engineering, manufacturing constraints, and commercial viability. Organic liquid, aqueous, ionic-liquid, concentrated, inorganic solid, polymer, and composite electrolytes are compared using transport, stability, processing, and interface criteria. The principal cathode and anode families are then evaluated in terms of practical voltage, reversible capacity, cycling stability, raw-material exposure, manufacturability, and end-of-life implications. A distinctive contribution of this work is the explicit separation of thermodynamic predictions, laboratory measurements, prototype demonstrations, and company-reported targets, together with design rules that connect electrolyte chemistry to cell-level performance. Sodium-ion batteries are unlikely to replace lithium-ion batteries universally, but they can occupy a strategic role where cost, safety, abundance, supply-chain resilience, and circularity outweigh maximum energy density.
Li-ion batteries (LIBs) are seeing increasingly widespread adoption across consumer electronics, electric vehicles, and grid-scale energy storage systems, yet their susceptibility to thermal runaway remains a concern. This study evaluates ethoxy (pentafluoro) cyclotriphosphazene (PFPN) as an electrolyte additive to reduce electrolyte flammability and thermal stability without significantly compromising electrochemical performance. Electrolyte flammability was quantified using self-extinguishing time (SET) measurements, which revealed that PFPN significantly suppresses combustion. At 4 wt% PFPN, 60% of electrolyte samples failed to ignite despite extended ignition exposure, and the average SET decreased from 51.15 s g−1 to 34.60 s g−1. Differential scanning calorimetry (DSC) further demonstrated improved thermal stability, with the onset of solvent decomposition delayed by ~30 °C at 4 wt% PFPN. Ionic conductivity modestly decreases (14%, from 8.13 to 6.97 mS cm−1 at 4 wt% PFPN). Electrochemical testing showed negligible impact on battery performance. Graphite||Li and NMC811||Li half-cells containing PFPN exhibited comparable capacity retention to baseline cells. NMC811||graphite pouch cells were used to further evaluate extended cycling and rate capability; PFPN-containing cells demonstrated similar capacities even after prolonged cycling and high-rate operation. Overall, PFPN provides effective flame retardance at 4 wt% while maintaining electrochemical compatibility, making it a promising additive for enhancing thermal stability of LIB electrolytes.
Precise estimation of lithium-ion battery State of Health (SOH) is highly demanded for reliable battery management systems, lifetime prediction, and safety assurance in electric vehicle and energy storage applications. Traditional data-driven approaches such as multilayer perceptron (MLP) often suffer from poor generalization and may produce non-physical degradation trends due to the absence of domain knowledge constraints. To address these limitations, this work proposes a monotonic Physics-Informed Residual MLP neural network framework for SOH estimation using the NASA battery dataset (B0005, B0006, B0007, and B0018). The proposed model incorporates a physics-based monotonic degradation constraint by penalizing positive gradients of SOH with respect to cycle index, thereby enforcing physically consistent capacity fade behavior. A loss function is employed to improve robustness and enhance late-cycle learning. Experimental results demonstrate that the proposed approach achieves an RMSE of 0.0287, MAE of 0.0181, and MAPE of 2.69%, indicating accurate and stable SOH prediction across multiple degradation patterns. The use of physics-informed constraints markedly enhances deterioration consistency and diminishes overfitting relative to solely data-driven models. The proposed structure offers a faithful solution for State of Health estimation in practical battery management systems.
Accurate state of health (SOH) estimation of lithium-ion batteries is critical to the reliability of electric vehicles. However, under dynamic operating conditions, conventional feature extraction based on fixed time window often exhibits poor generalization, as it fails to account for the multi-timescale parameter couplings inherent in the non-stationary voltage responses. To address this issue, this paper proposes a Bayesian Adaptive Time Window Optimization (BATWO) framework for feature extraction in battery SOH estimation. Within this framework, the time window length is treated as a learnable structural parameter and is adaptively optimized via Bayesian optimization to identify the most informative observation timescale for extracting degradation-sensitive statistical features under given operating conditions. Evaluations on a cycle-aging dataset containing 69 lithium-ion battery samples subjected to distinct dynamic operating profiles show that the optimal time window lengths vary significantly, ranging from 500 s to 27,630 s. The BATWO framework achieves an average root-mean-square error (RMSE) of 2.07% and a mean absolute error (MAE) of 1.45%, outperforming the best fixed time window strategy by reducing the RMSE and MAE by 2.35% and 2.68%, respectively. Moreover, compared with LSTM- and Transformer-based models without feature extraction, the BATWO framework reduces training time by over 97%. These results highlight the superior generalization capability and computational efficiency of the BATWO framework, demonstrating its great potential for practical deployment in battery management system.
There is a large and rapidly growing stock of retired power batteries from new energy vehicles in China. Unregulated informal recycling and improper disposal of these waste batteries trigger serious environmental hazards. Though a batch of regulatory policies on battery recycling have been released in recent years, the power battery recycling sector still faces prominent governance bottlenecks, especially ambiguous responsibility division and poor implementability under the entrusted recycling mode. To fill the existing research gap regarding tripartite interest conflicts and pricing mechanisms in entrusted recycling, this paper constructs a three-party evolutionary game model covering power battery producers, recyclers and government regulators. Two pricing models are further developed to distinguish producer self-operated recycling and third-party entrusted recycling channels. Numerical simulation is adopted to investigate multi-stakeholder interest contradictions, dynamic evolutionary trajectories and equilibrium stability of the recycling system, and the influences of subsidy intensity, supervision intensity and recycling cost on participants’ strategic choices are quantitatively analyzed. The research results demonstrate that inadequate government supervision and insufficient economic returns for formal recyclers serve as the primary obstacles hindering the effective deployment of entrusted recycling. An inherent and reasonable price gap exists between self-operated and entrusted recycling modes. Essentially, the price differential of standardized entrusted recycling represents the profit margin conceded by producers to recyclers instead of direct financial subsidies. To solve existing industry problems, this study proposes targeted recommendations for tripartite collaboration. The government should refine the regulatory framework of Extended Producer Responsibility and adopt differentiated reward and penalty mechanisms. Producers are expected to standardize entrusted recycling management and formulate a scientific pricing range for retired batteries. Recyclers ought to advance recycling technologies and maintain standardized operations. Collective efforts from all stakeholders can facilitate the long-term sustainability of the closed-loop recycling system for retired power batteries.
This paper proposes an SOH estimation method that fuses mechanical–electrical–thermal multi-modal features by introducing expansion force monitoring. Aging tests on 16 prismatic 530 Ah LiFePO4 batteries from two brands are conducted at 25 and 45 °C. Each full cycle is divided into charge, post-charge rest, discharge, and post-discharge rest, with SOH defined by the capacity ratio. From cycle-level data, 37 candidate features are extracted and cleaned using local median and median absolute deviation. Using only training cells, Spearman correlation eliminates highly redundant features, and 12 key features are retained via internal validation. Under 4-fold cross-validation with complete battery grouping, Random Forest, XGBoost, LightGBM, and LSTM are compared. LightGBM achieves the best performance with an average MAE of 0.0032, RMSE of 0.0037, and R2 of 91.36%. Ablation shows multi-modal fusion outperforms single-type features; five-category fused features reduce RMSE by ~75.57% versus electrical-only features. Removing expansion force features increases RMSE to 0.0064 and drops R2 to 75.66%. These findings confirm that expansion force supplies critical mechanical degradation information, significantly improving SOH estimation for large-capacity energy storage batteries.
The Extended Kalman Filter (EKF) serves as a widely utilized approach to evaluate the state of charge (SOC) of energy storage batteries. However, the conventional EKF is commonly adversely affected by ambient temperature variations, uncertain noise matrices, and inaccurate parameter estimation in practice. Therefore, hybrid elephant herding and golden eagle optimization based EKF (HEGO) is introduced to enhance the precision and effectiveness of battery SOC estimation. The local contraction capability of elephant herding optimization is utilized to narrow the search range within a predefined search space and accurately locate the region of the optimal solution. Within the narrowed search range provided by EHO, golden eagle optimization (GEO) is then employed to accurately identify the noise matrix and equivalent circuit parameters appropriate for the current state, thereby increasing the precision and resilience of SOC estimations against environmental disturbances. Data for an 18650-battery evaluated with the Federal Urban Driving Schedule (FUDS), Dynamic Stress Test (DST), and Hybrid Pulse Power Characterization (HPPC) conditions were collected using an experimental platform, and the proposed algorithm was experimentally validated. The results demonstrate that, across different temperatures and operating conditions, the proposed algorithm consistently achieves optimal performance, with a mean absolute error below 0.7% and strong generalization, thereby providing stable and reliable technical support for battery SOC estimation.
Accurate state-of-health (SOH) estimation is essential for the reliable operation of lithium-ion batteries. However, predicting SOH for previously unseen batteries remains challenging because degradation behavior varies among cells. This study proposes a multi-source feature-learning framework that combines electrochemical impedance spectroscopy (EIS), discharge-profile, and aging-related information for cross-battery SOH estimation. EIS and discharge data from 34 batteries in the National Aeronautics and Space Administration (NASA) battery aging dataset are processed to construct 1830 matched impedance–SOH samples. A total of 101 features are extracted from raw and rectified impedance spectra, resampled impedance points, NASA-provided impedance parameters, discharge profiles, and cycle-related information. Random Forest (RF), Extra Trees (ET), Gradient Boosting (GB), Histogram Gradient Boosting (HGB), and Extreme Gradient Boosting (XGBoost) models are evaluated using random sample splitting and strict battery-wise validation. Under random validation, GB achieves the best performance, with a coefficient of determination (R2) of 0.9788. Under strict battery-wise validation, ET achieves a mean absolute error (MAE) of 4.9131 percentage points, a root mean square error (RMSE) of 7.3664 percentage points, and an R2 of 0.7855. The performance difference between the two validation strategies demonstrates the importance of battery-grouped evaluation when assessing generalization to unseen cells. Overall, the results indicate that combining impedance- and discharge-derived information provides a promising basis for cross-battery SOH estimation, although further leakage-controlled validation across broader operating conditions is required.
